HARVARD UNIVERSITY LIBRARY OF THE Museum of Comparative Zoology S- NAR A omrence} hance toxin 7 ‘? A pan ‘stag led 5 aa Ae WiLISs COMP. 27 OO! g PA rs a pe Line apy : f prs i UNIVERSITY OF KANSAS JUN 16 1965 b JALAL SCIENCE BULLETIN.» UNIVERSITY OF KANSAS PUBLICATIONS University of Kansas Science Bulletin - Vol. XLV June 7, 1965 Lawrence, Kansas ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas University Quarterly) is issued in part at irregular inter- vals. Each volume contains 300 to 700 pages of reading matter, with necessary illustrations. Exchanges with other institutions and learned societies everywhere are solicited. All exchanges should be addressed to THE UNIVERSITY OF Kansas SCIENCE BULLETIN, LIBRARY OF THE UNIVERSITY OF KANSAS, LAWRENCE, KAn. PUBLICATION DATES The actual date of publication (i. e., mailing date) of many of the volumes of the University of Kansas Science Bulletin differs so markedly from the dates bourne on the covers of the publication or on the covers of the separata that it seems wise to offer a corrected list showing the mailing date. The editor has been unable to verify mailing dates earlier than 1932. Separata were issued at the same time as the whole volume. Vol. XX—October I, 1982. Vol. XXXIV, Pt. I—Oct. 1, 1951. Vol. XXI—November 27, 1934. Pt. II—Feb. 15, 1952. Vol. XXII—November 15,1935. Vol. XXXV,Pt. I—July 1, 1952. Vol. XXIII—August 15, 1936. Pt. Il—Sept. 10, 1958. Vol. XXIV—February 16, 1938. Pt. III—Nov. 20, 1953. Vol. XXV—July 10, 1989. Vol. XXXVI, Pt. I—June 1, 1954. Vol. XXVI—November 27, 1940. Pt. [I—July 15, 1954. Vol. XXVII, Pt. I—Dec. 30,1941. Vol. XXXVII, Pt. I—Oct. 15, 1955. Vol. XXVIII, Pt. I—May 15, 1942. Pt. I]—June 29, 1956. Pt. II—Nov. 12, 1942. Vol. XXXVIII, Pt. I—Dec. 20, 1956. Vol. XXIX, Pt. July 15, 1948. Pt. Ii—March 2, 1958. Pt. 1I—Oct. 15, 1948. Vol. XXXIX—Nov. 18, 1958. Vol. XXX, Pt. 1—June 12, 1944. Vol. XL—April 20, 1960. Pt. II—June 15,1945. Vol. XLI—Dec. 23, 1960. Vol. XXXI, Pt. I—May 1, 1946. Vol. XLI—Dec. 29, 1961. Pt. II—Nov. J, 1947. Vol. XLII—Supplement to, Vol. XXXII—Nov. 25, 1948. June 28, 1962. Vol. XXXIII, Pt. I—April 20, 1949. Vol. XLITI—Aug. 20, 1962. Pt. 1I—March 20, 1950. Vol. XLIV—Sept. 1, 1963. J OGTTROV A shat gee ey Nt Rurus H. THomepson Editorial Board........ Rurus H. Tuompson, Chairman KENNETH ARMITAGE CHARLES MICHENER Pau ROOFE WortHit H. Horr Davip PARETSKY Parke H. Wooparp, Secretary UNIVERSITY OF KANSAS SCIENCE BULLETIN DEVOTED TO THE PUBLICATION OF THE RESULTS OF RESEARCH BY MEMBERS OF THE UNIVERSITY OF KANSAS VOLUME XLV UNIVERSITY OF KANSAS PUBLICATIONS LAWRENCE, JUNE 7, 1965 PRINTED BY HARRY (BUD) TIMBERLAKE, STATE PRINTER TOPEKA, KANSAS eo 30-1367 Contents of Volume XLV NUMBER 1. Coefficients of Correlation and Distance in Numerical Maxononiyace ey F. James Rohlf and Robert R. Sokal, 2. Generic Relationships of the Family Pyemotidae (Aca- rinasrombiditormes)\se. a ee Earle A. Cross, 3. A revision of the Banded Subgenera of Nomia in America (Hymenoptera: Halictidae) ...... David W. Ribble, 4. A Taxonomic Study of the Subgenus Ladona (Odonata: Iibellulidae yy a eee Billy L. Bennefield, 5. The Genus Ptilomera Amyot and Serville (Gerridae: Hemiptera ). Herbert B. Hungerford and Ryuichi Matsuda, 6. A Morphological Study of the Autonomic Nervous System of the Opossum, Didelphis marsupialis.. Lily S. Feng, 7. The Structure and Innervation of the Venom Glands in the Tail of Salamanders (Ambystoma). Grant A. Mason, Jr., James L. Hall, Paul G. Roofe, 8. The Sympathetic and Parasympathetic Nerve Fibers in the Renal Cortex of the Cat. Phyllis D. Zwarych, Paul G. Roofe, 9. The Serpents of Thailand and Adjacent Waters. Edward H. Taylor, 10. A Chemical and Limnological Study of Lake Vanda, Vic- toria Land, Antarctica. Ernest E. Angino, Kenneth B. Armitage and Jerry C. Tash, 11. Stelastellara parvula, a new genus of unknown affinity from the American Carboniferous .. Robert W. Baxter, PAGE On — “I 557 587 609 1097 1119 THE UNIVERSITY OF KANSAS SCIENCE BULLETIN VoL. XLV] JUNE 7, 1965 [No. 1 Coefficients of Correlation and Distance in Numerical Taxonomy ? ? BY F. JaMEs Rour ® and RosBert R. SoKat * Department of Entomology The University of Kansas, Lawrence Asstract: This is an investigation of the properties of two coefficients of similarity in numerical taxonomy: correlation (Michener and Sokal, 1957) and distance (Sokal, 1961). The study is based upon analyses of 97 species of bees by Michener and Sokal (1957) and 48 species of mosquitoes by Rohlf (1962). Characters in these studies generally possessed several linearly ordered states. The effect of standardization of characters (i. e., transformation of the character state codes to achieve a mean of zero and a variance of unity) is slight upon the distance coefficient. Correlation coefficients are affected to a greater degree and there is a definite nonlinear relationship between correlation co- efficients based on standardized and on unstandardized characters. Compari- sons between coefficients of correlation and distance based upon standardized characters do not as yet permit conclusions on the relative merits of the two, except in special cases. Biological and mathematical interpretations of these coefficients are discussed. The theoretical frequency distributions of these coefficients are given and compared with the observed distributions. The properties of several alternative distance coefficients are briefly examined. A summary of recommendations for carrying out numerical taxonomic studies is given, based on the present state of our knowledge. 1. Contribution number 1175 from the Department of Entomology, University of Kansas, and paper No. 4 in a series entitled Experiments in Numerical Taxonomy, of which Sokal (1961), Rohlf and Sokal (1962), and Sokal and Rohlf (1962) are the first three papers, respectively. 2. This investigation was supported in part by a research grant from the National Insti- tute of Allergy and Infections Diseases, U. S. Public Health Service. 8. Present address: Department of Biological Sciences, University of California, Santa Barbara. Parts of this investigation were carried out during the tenure of a predoctoral fellowship from the Division of General Medical Sciences, United States Public Health Service, by Rohlf. 4. We are indebted to Mrs. Julie C. Sokal, Richard C. Rinkel and Paul A. Thomas for computational assistance. Mrs. Maxine L. Howe helped with the clerical work and Mrs. Julie C. Sokal prepared the illustrations. 4 THE UNIVERSITY SCIENCE BULLETIN I. INTRODUCTION The present paper investigates the effect of standardization of characters on correlation and distance coefficients in numerical taxonomy. These studies are carried out using the data on 97 species of bees by Michener and Sokal (1957) and 48 species of mosquitoes analyzed by Rohlf (1962). As numerical taxonomy develops and its application spreads it is important to examine theoretically and empirically the validity of its assumptions and techniques. Three types of similarity coeffi- cients have been proposed in numerical taxonomy: association coefficients, correlation coefficients, and distance coefficients. The latter two have been most frequently used with higher organisms owing to the nature of the characters employed in taxonomic work on these groups. No critical comparison has so far been made be- tween these two types of coefficients. The present paper, while not definitive, will provide the first comparative study of these coeffi- cients. As early as 1957 Michener and Sokal recognized that in order to overcome the distortions in correlation coefficients due to different numbers of states per character it was desirable to standardize the characters to give them a common mean and unit variance. How- ever, the limitations of their computational equipment at the time did not permit them to carry out such computations. The effects of standardization on Michener and Sokal’s data are described here and compared with the earlier results. In his study of the mosquito genus Aedes, Rohlf (1962) employed both standardized and un- standardized character state codes. The comparisons are presented below. Since the principles and procedures of numerical taxonomy have been adequately detailed elsewhere (Sokal, 1960; Sneath and Sokal, 1962; Sokal and Sneath, 1963) no further discussion of these is made here. Interested readers are referred to these papers. II. MATERIALS The original bee data from Michener and Sokal (1957) were repunched on a new set of IBM cards in order to permit processing by newly-developed techniques for the IBM 650 computer. The general correlational and clustering procedures were as described by Sokal and Michener (1958). Detailed modifications will be described below as needed. The programs for processing these data were largely developed by one of us (FJR). The 48 species NUMERICAL TAXONOMY INVESTIGATIONS 5 of the genus Aedes were employed in a numerical taxonomic study by Rohlf (1962). They belong to eight subgenera. Both larval and adult characters were studied and the correlations and distances were computed between species based on larval, adult and both sets of characters. The computations presented here were by- products of an investigation examining the congruence of larval and adult classifications (Rohlf, 1963). In the bee study 122 characters were employed, while the mosquito study was based upon 148 characters. Lists of the species and characters studied and biologi- cal information on them are furnished in the original sources already referred to. II. STANDARDIZATION OF CHARACTERS Standardization seems desirable since the coding of the various character states is arbitrary and consequently the number of states and the units of measurement vary from character to character. In order to pool the information from different characters, it is desirable to transform all characters in a given study to a comparable scale. Early steps in this direction were taken by Smirnov (1925) and by Haltenorth (1937). The former standardized deviations of taxa from the means of their groups (by division by the mean), while the latter expressed all measures (of the skulls of cats) as percent- ages of the basal lengths of these skulls. Cain and Harrison (1958) dealt with the problem of standardiza- tion by using the maximum expression of a character as 100 percent and arbitrarily grading all other expressions of the character as some portion of this total percentage. They called the transformed char- acter state codes “reduced values.” Since they used zero when a character was absent (or minimal), the effect of their transforma- tion was to give all of the characters an equal range. There are two difficulties with this procedure. First, the admission of an OTU (operational taxonomic unit—the lowest taxonomic unit being classified—see Sokal and Sneath, 1963, for a more detailed discus- sion) with a character state scored beyond the previous 100 percent level would necessitate the recalculation of all of the “reduced” state codes for the character; hence, the degree of similarity be- tween the OTU’s previously studied would be affected. The second is that the reduced character state codes are not readily amenable to rigorous statistical methods. A procedure proposed for use in numerical taxonomy by Sokal (1961) is standardization of the characters (i. e., rows of the data matrix). To do so we transform each character to obtain a mean 6 THE UNIVERSITY SCIENCE BULLETIN of zero and a variance of unity. This is easily done by the follow- ing transformation: Kee X' i, = ————_ Si where Xj; is the character state value for OTU j on character i, X; is the mean of character i, s; is the standard deviation of character i, and X’;; is the standardized character state code. The inclusion of additional OTU’s to a numerical taxonomic study with standardized characters will not have much effect on the means and standard deviations of the characters unless the new character states are very extreme and/or the previous study was based on very few OTU’s. Hence, it usually will not be necessary to re-standardize the characters every time a new OTU is added to the study. Another desirable feature is that the relationships among the previously studied OTU’s will be affected little by the addition of other OTU’s. The procedure to use when a new OTU is added, is to express its character state codes in terms of character means and standard deviations based on the previously studied OTU’s. IV. SIMILARITY COEFFICIENTS The selection of the appropriate coefficient depends upon the assumptions made about the characters used in the study and the manner in which they were recorded (Rohlf, 1962; Sokal and Sneath, 1963). If the states of each character used could not be linearly ordered a coefficient of association would be the most appropriate (however, in two-state characters, whether linearly ordered or not, all three types of coefficients can be used). A common coefficient of association is the simple matching coefficient which is the number of characters for which two OTU’s match divided by the number of characters considered (Zubin, 1938; Cattell, 1949; and Sokal and Michener, 1958). This type of coefficient has been often used in psychology and ecology. Sneath (1957) and Rogers and Tanimoto (1960) show taxonomic applications of various association coefh- cients. In the present study only correlation and distance co- efficients are considered. a. Correlation Coefficients The only correlation coefficient that has so far been employed in numerical taxonomy is the Pearson product-moment correlation co- efficient. It was introduced to numerical taxonomy by Michener and Sokal (1957) and Sokal and Michener (1958). Its formula is too well-known to need repetition here. NUMERICAL TAXONOMY INVESTIGATIONS ie Correlation coefficients can be interpreted in several ways. Their common interpretation is as a measure of covariation. If two vari- ables vary in such a way that as one increases in value the other one does likewise and conversely, as the first one decreases the second one follows, then these variables covary and are said to be positively correlated. In numerical taxonomy this means that two OTU’s are positively and highly correlated if changes in character state score from character to character in one OTU are paralleled in the other. A useful interpretation can be made through geometry when standardized characters are used. Figure 1 will explain this in a if Z Fic. 1.—Two dimensional representation of three OTU’s plotted in a three dimensional space. The OTU’s A, B, and C are plotted with respect to three character axes X, Y, Z defining a three dimensional space. Distances between OTU’s are identified by A, Correlations between OTU’s are functions of the angle between the lines connecting them to the origin, O. For further explana- tion see text. 8 THE UNIVERSITY SCIENCE BULLETIN simplified model based on only three characters. These characters are called X, Y and Z and represented by the three co-ordinate axes of the figure. Each OTU (three are shown in the present diagram but, of course, any number could be used) is plotted in the three- dimensional figure according to the scores of its character state codes for the three characters. Each OTU therefore assumes a fixed position in the three-dimensional space. The angle between the lines connecting any two OTU’s to the origin is a measure of the similarity between them, since in general the closer the OTU’s are to each other in the three dimensional space the smaller this angle will be. The correlation coefficient can be interpreted as the cosine of this angle (which will be 1 for an angle of 0°, 0 for an angle of 90° and -1 for an angle of 180°). This model is not strictly true since the columns of the data matrix (the OTU’s) do not have means exactly equal to zero but when standardized char- acters are used the model is accurate enough to permit one to visualize the relationships. In the present study it was found that the observed means of the OTU’s, after the characters had been standardized, were very close to zero. The importance of standardization when correlation coefficients are used can be seen from the fictitious example in Table 1. OTU’s TaBLE 1.—Hypothetical Data to Demonstrate Effect of Sandardization of Characters. | OTU’s CHARACTERS | Possible state codes — 1 2, 1 1.2 1 2 2 12 3 3 1 3 23 4 1 4 4 12354805 5 1 5 1 34 5G 1 6 1 and 2 exhibit maximum differences for the given range of char- acter state codes, yet the correlation between OTU’s | and 2 is only —0.696, and not -1.000 as one might expect. When the characters are transformed, by equalizing their range or by standardization or even by simply giving all of them the same mean, the correlation between the two OTU’s is -1.00. It would seem that with unstandardized data, it is impossible to get very NUMERICAL TAXONOMY INVESTIGATIONS 9 a hi / / \ \ \ \ \ 1000 ’ \ \ \ \ 500 2000 500 Fic. 2.—Observed and expected frequency distributions of correlation co- efficients from the bee study of Michener and Sokal (1957). a. Upper graph shows correlation coefficients based on characters that had not been standard- ized. b. Lower graph shows correlations based upon standardized characters. Abscissas give magnitude of correlation coefficient, r, while ordinates indicate frequencies of correlation coefficients. Solid lines show observed frequencies, dashed lines show expected frequencies on the basis of the observed mean in graph a and of a mean of zero in graph b. high negative correlations; it is, therefore, probably unlikely that in any actual study many negative correlations would be found at all. The above appears to explain why Michener and Sokal (1957) found virtually no negative correlations in their study, based upon unstandardized data (Figure 2a), while Rohlf (1962) found no negative correlations among his coefficients based on unstandard- ized data (Figure 3a). When the correlations were recomputed based on stardardized characters, the average correlation among the bees decreased from 0.418 to -0.009 (see Figure 2b), while 10 Tue University SCIENCE BULLETIN 500 400 300 200 100 300 200 100 Fic. 3—Observed and expected frequency distribution of correlation co- efficients from the mosquito study of Rohlf (1962). Explanation as in Figure 2. in the Aedes data a corresponding shift in the average correlation from 0.703 to —0.018 took place (Fig. 3b). Standardization of the character state codes raised the standard deviations slightly in both studies, from 0.185 to 0.253 and from 0.100 to 0.157 in bees and mosquitoes, respectively. From an in- spection of the frequency distributions in Figs. 2 and 3 it can be NUMERICAL TAXONOMY INVESTIGATIONS VI seen that the standardization of the characters affects both the position of the frequency distribution along the abscissa and also the shape of the distributions. It will be noted that the average correlation for each matrix of coefficients based on standardized characters approached zero very closely and cannot in fact be shown to be significantly different from zero. We believe the expected value of this average (based on stardardized characters) to be zero, although we have not dem- onstrated this mathematically. It is of interest to study the observed frequency distributions of the product-moment correlation coefficients in relation to their theo- retical distributions. These were calculated using an expected value of zero for the correlations based on standardized characters and on an expected value equal to the observed mean correlation for the correlations based on unstandardized characters. The expected distribution of the correlation coefficient has been tabled by David (1938), but Rohlf (1962) found it more convenient to approximate the distribution using z’s than to interpolate in her tables, since a simple linear interpolation was found not to be sufficiently accurate. For n greater than 25 (true of all our cases), the z transformation gives results very close to the actual distribution (David, 1938). Figures 4 and 5 show two-way frequency distributions of the correlations based on unstandardized and standardized characters for each pair of species. The frequencies in each cell of the two-way frequency distribution are represented by black circles. The size of each circle indicates the magnitude of the frequencies it repre- sents according to a key shown in the upper left corner of each figure. The general shapes of these two frequency distributions are quite similar. The relationship is clearly nonlinear and appears constrained by the upper limit of the correlation based on unstand- ardized characters. When correlation coefficients between the two types of r’s are calculated for the data on which these figures are based we find these to be 0.89 and 0.63 for the bees and mosquitoes, respectively. b. Distance Coefficients A distance coefficient expresses the dissimilarity between two OTU’s as a geometric distance between the two points (correspond- ing to the two OTU’s) in an n-dimensional Euclidian space defined by the n characters as co-ordinate axes. Sokal (1961) discusses the geometric interpretation of a distance coefficient and its relation to other similar indices. A distance coefficient can be easily calculated Y THE UNIVERSITY SCIENCE BULLETIN standardized ) r (characters 1.0 -.2 6) 2 4 6 8 r (characters not standardized ) Fic. 4.—Two-way frequency distribution of the correlation coefficients based on unstandardized and standardized characters for each pair of species in the bee study by Michener and Sokal (1957). The abscissa and ordinate indicate the magnitude of the correlation coefficient. The frequencies in each cell of the two-way frequency distribution are represented by solid circles. The size of each circle indicates the magnitude of the frequency according to the key shown in the upper left corner of the graph. The graph is based on 4,656 correlation coefficients. The correlation coefficient between the two variables is 0.89. (assuming orthogonal co-ordinate axes) by an extension of the Pythagorean Theorem to an n-dimensional space. The taxonomic distance, between OTU’s j and k, 4;,, is defined as Vo A ix — | Sut — xa) | where X,,; is the state code of OTU j for character i and n is the number of characters in the study. The quantity 4;,? was called Six” by Sokal (1961). We are changing symbolism here (in agree- ment with Sokal and Sneath, 1963) in order to conform to the conventional statistical usage of employing lower case Greek letters for parameters. Since we are reserving the lower case English d for average distance as specified below and since D is used for Ma- NUMERICAL TAXONOMY INVESTIGATIONS 18 r (characters standardized ) 5 ( £ ol to = ° = ip ol ‘s on fo) o on Dee 0 e ® @ i) eee b @ ee @ee ee a e eo e e@ @ Been ee e e @ @ S$ e © @oeee e ao Pa} e @e@ee«e#eie io S Ss @eeee e eo ee @e @ @ ee e@© @ e > 2} “re @oe@eee @e 3° ©-e«@@e-e- ° ° : -©00@@e-- . Sc | -@0e@O@@eer-: . ; @Q a -©00000 a e e@ Oe ee: yy -@0 1. In such a case distance would have no information not con- tained in the correlation coefficient. Near the upper limit for cor- relations (lower limit for distances) the distance scale would be expanded relative to that of r. It is this tail of the scale which generally is of more interest in numerical taxonomy. (1-rjx). Since OTU’s in actual practice are not standardized and are un- likely to be equidistant from the origin, the distance coefficient does convey independent information. Lengths of the lines between the OTU’s and the origin are a measure of “size” of the OTU as expressed through its characters. This is a rather vague term be- cause some characters such as color will not express size dimensions in the conventional meaning of the word. NUMERICAL TAXONOMY INVESTIGATIONS 15 Pe) 5 1.0 Ls 2.0 d Fic. 6.—Graph of the distance coefficient, d, as a function of the correlation coefficient, 7, on the assumption that the OTU’s are standardized. The mathe- matical expression is shown in the text. The abscissa shows d and the ordinate r contrary to ordinary conventions in order to conform with the arrangement in Figures 10, 11, and 12. It is of interest to examine the frequency distributions of the observed distance coefficients in the bee and mosquito data and to compare them with the theoretical distribution that one would expect if the distance coefficients were all drawn from the same statistical population, i. e., if the true distances between all pairs of OTU’s were equal and the observed distribution of distances about their mean were due to chance alone. The observed frequency distributions of distances between pairs of species in bees and mosquitoes are shown in Figure 7 and 8, respectively. Distances were computed from standardized characters in both studies and also from unstandardized characters in the mosquito study. The expected frequency distributions are superimposed on the observed frequency distributions based on standardized characters only. In order to calculate these expected frequency distributions, it was assumed that the observations (the n standardized characters ) 16 THe UNIVERSITY SCIENCE BULLETIN 2500 2000 1500 1000 500 Fic. 7.—Observed and expected frequency distributions of distance coeffi- cients based on standardized characters from the bee study of Michener and Sokal (1957). Abscissa shows distance scale, ordinate indicates frequencies. The solid line indicates observed frequencies, the dashed line gives expected frequencies based on the assumptions stated in the test. were independent, normally distributed with a mean of zero and with unit variance. On these assumptions the variance of a differ- ence between two variables is distributed as 2x? (with one de- gree of freedom). Hence, average distances are distributed as V2xn7/n (where x,7 has the y?-distribution with n degrees of freedom). The easiest way to prepare the frequency distribution of such a variable is to use a transformation to another distribution which has already been adequately tabled. A suitable transforma- tion to use would be: This is based upon an equation for calculating the distribution of x2 for large degrees of freedom (n). The transformed variable wu is then expected to be normally distributed with a mean of zero and NUMERICAL TAXONOMY INVESTIGATIONS ik7e 400 a 300 100 400 300 f 200 100 6 .8 1.0 1.2 1.4 L6 18 2.0 22 2.4 Fic. 8.—Observed and expected frequency distributions of distance coeffi- cients from the mosquito study of Rohlf (1962). a. Upper graph shows ob- served distribution for distances based on characters that had not been standardized. b. Lower graph shows distances based on standardized char- acters. Abscissas give magnitude of the distances, ordinates indicate frequen- cies. Solid lines represent observed frequencies, dashed line represents the expected distribution based on the assumptions stated in the text. with unit variance (Kendall and Stuart, 1958, p. 8373). Within the range of the observed d’s, and for large n, this is practically a linear transformation. Under the above assumptions the expected value of d has been shown by Rohlf (1962) to be (for n even) (n— 1)! | E(d) = —————— ie: 2—1367 18 THE UNIVERSITY SCIENCE BULLETIN Using Stirling’s formula, this reduces to: MWe 1 Bee walls eee n (= 74) e lim aie ee | ave For n = 122 and n = 148, the numbers of characters in the bee and Aedes studies, the expected values for d are 1.411 and 1.412, repec- tively. The observed means of the distances are 1.391 and 1.383, respectively, and thus are smaller than the expected values. The expected variance of d is: Caz = 2 [E(d)]? Hence: which approaches zero as n tends to infinity. For n= 122 and 148, oa? = 0.009, and 0.006, which are much smaller than the observed variances, 0.0660 and 0.0864, respectively. Sokal (1961) stated that the expected value of d would be 1.12 [1.128] for any n. He based this on the expected value of the absolute difference between two standardized normal variates. However, d is the square root of the mean squared difference and not the arithmetic mean of absolute differences which would have an expected value of approximately 1.128, since the expected value of the absolute difference is 2c/\/z (Kendall and Stuart, 1958). Figure 9 shows the two-way frequency distribution of the dis- tances based on unstandardized and standardized characters for each pair of mosquito species. Each point represents a pair of species. The main effect of standardization was to raise both the mean and the standard deviation of the distance coefficient. This can be explained by the fact that most of the characters used had standard deviations less than unity. The scattergram looks con- siderably more linear than did the similar scattergrams for corre- lation coefficients (Figs. 4 and 5). This close relation between distances based on standardized and unstandardized characters (r= 0.91) can be explained simply, since changing the origin of the character axes does not change the distance between OTU’s. Differences in the values of the distances after standardization are functions of changes in the variances of the characters upon stand- ardization. These differences are expected to be quite minor and were definitely so in the mosquito study. From these results it would seem that for purposes of numerical taxonomy there is not much difference between standardized and unstandardized dis- NUMERICAL TAXONOMY INVESTIGATIONS 19 a © r (characters not standardized ) “A 6 8 LO 12 1.4 16 d ( characters not standardized ) Fic. 9.—Two-way frequency distribution of the distance coefficients based on unstandardized and standardized characters for each pair of species in the mosquito study by Rohlf (1962). Explanations of this figure and scale of solid circles as in Figure 5. tances. As would be expected, the dendrograms are also rather similar. However, on general principles, distances based on stand- ardized characters are to be preferred. To use unstandardized characters is equivalent to weighting the characters in proportion to their observed standard deviations. It is also of interest to note that even though the largest standard deviation was about one hundred times as large as the smallest, the effect on over-all relation- ships was very small. This shows that in order to have a strong effect, a system for weighting characters would have to employ an extreme set of weights. Some further theoretical observations on the distance coefficient and on some other coefficients follow below. The distance coefh- cient of Sokal (1961) can be easily transformed into the form of a correlation coefficient (7p, the coefficient of pattern similarity of Cattell, 1949), but no additional information is gained by so doing. x2n(.50) — nd? ; Serres) ea aR x2n(.50) + nd?i; where n is the number of characters and x,7(.50) is the median of 20, THE UNIVERSITY SCIENCE BULLETIN a x°-distribution with n degrees of freedom. This coefficient must be used on standardized characters. Within the range of distance coefficients observed in numerical taxonomic studies, and for large n, there is practically a linear relationship between d and r,; hence, dendrograms based on the different indices would be essentially identical. Penrose (1952) shows how the distance coefficient can be parti- tioned into “size” and “shape” components which are supposed to give the relative contributions of these factors to the difference be- tween two OTU’s j and k. These components are: 9 n S (Xi; — Xix) SAG aO)2 == 0 — (Site) 4 = rh n with symbols defined as before. As can be seen, (Shape)? is the variance of the differences between OTU’s j and k and will be zero, i.e., indicating identity in “shape” only if the difference between the two OTU’s is constant for all of the characters. Hence, this is a measure of similarity in over-all shape only if the character state codes for an OTU are all equal or if the OTU’s are identical in both size and shape. For example, consider Table 2 which is based on hypothetical data. OTU’s 3 and 4 would have identical shapes by Penrose’s “shape” coefficient, but 1 and 2 would not, even though the character state codes of OTU 2 are simply twice that of OTU 1. TABLE 2.—Matrix of Hypothetical Data. CHARACTERS OTU’s 1 2, 3 4 5 1 1 2 2 4 2 2, 2, 4 2, 4 3 3 3 6 2 4 4 OTU’s 1 and 5 would also be indicated to be identical in shape despite the fact that they are not of the same proportions. The NUMERICAL TAXONOMY INVESTIGATIONS Pall product-moment correlations correctly indicate that OTU’s 1 and 2, but also 5, have the same “shape.” The correlation between OTU’s 3 and 4 cannot be calculated because their variances are zero. This would indicate that in most cases the correlation coefficient is an adequate measure of similarity in “shape” since (1) it is very un- likely that an OTU would have identical character state codes for all of its characters, and (2) using standardized characters the means of the OTU’s will usually be close to zero; hence, the regression lines will pass close to the origin so that cases such as OTU 1 versus 5 are not likely to occur very often. _ The consideration of size and shape is an important one in nu- merical taxonomy. If most of the characters used in a study are measurements of various parts of an organism, and if the OTU’s differ much in over-all size (measured by the distance of an OTU from the origin), the distance coefficient will be more a measure of similarity in size than a measure of similarity in shape, the latter probably being more important for estimating over-all similarity in numerical taxonomy. A dendrogram constructed from such a study would group morphologically small OTU’s with small OTU’s and large OTU’s with large OTU’s. This, in most cases, would be very unsatisfactory for constructing a “general classification.” With data of this sort, correlation coefficients or special coding of the characters (see below) would seem preferable. This type of effect is also to be considered with meristic characters, although in this case there is less certainty as to what “size” represents. Rohlf (1962 ) reported such a “size” effect among his larval relationships. Dis- tances showed species 47 and 48 to be much more different from the rest of the species than was shown by the correlations. It ap- pears that there may have been a general “size” effect in these two species that caused many of their setae to become highly branched, the relative numbers staying about the same. In a sense, to say that there is a general size effect is mathematically the same as saying that the characters are, in general, highly correlated. From these observations it might be possible to suggest as a working guide, until a more rigorous mathematical treatment is made of this problem, that whenever the characters seem highly correlated, either correlation coefficients should be used or, before distances are used, the characters should be converted to ratios of some sort in order to eliminate as much correlation between characters as possible. When there is a mixture of types of characters, as is usually the case, it is difficult to say what the best procedure would be. 2a—1367 By) THe UNIVERSITY SCIENCE BULLETIN Cattell (1949) suggests using r, as a measure of similarity in shape regardless of level, where n S, X's; X ix N Sj Sk n are the number of characters, X’;; and X’;, are the standardized (by characters ) character state codes for OTU’s j and k, and s; and s; are the standard deviations of OTU’s j and k. However, since the values are not standardized by OTU’s this is not really a pure meas- ure of similarity in shape. There is still some “size” left in it. In order to get rid of this, one has only to use a correction factor to adjust for the differences in the means of the two OTU’s. When this is done, r,s = 7, the product-moment correlation coefficient. 25 @ e 20 @ie 7 @ N s } mo] c 5 7) e L5 @ oS 2 o <= 1S) Rs) 1.0 e e See eee al, 6 8 LO l.2 1.4 16 ad (characters not standardized ) Fic. 10.—Two-way frequency distribution of correlation and distance co- efficients for each pair of species in the mosquito study by Rohlf (1962). Both coefficients of similarity are based on characters that had not been standardized. Explanations of this figure and scale of the solid circles as in Figure 5. NUMERICAL TAXONOMY INVESTIGATIONS 23 c. Comparisons between correlation and distance coefficients As can be seen from the two-way frequency distribution of cor- relation and distance coefficients of Aedes based upon unstandard- ized data (Fig. 10) these two coefficients indicate very similar relationships among the species of mosquitoes. The correlation be- tween these two coefficients is —0.92 and an inspection of the figure reveals that the relationship is a linear one. The two-way frequency distribution for distance and correlation coefficients based on standardized data indicated lower correlations between the two coefficients in the bees (Fig. 11) and especially in the mosquitoes (Fig. 12). The observed correlations were —0.77 and -0.52, respectively. The species in the sub-genus Stegomyia and a few other species were much more different from the rest of the species in the study when measured by distance coefficients than when measured by correlation coefficients. This may have caused the much lower correlation in the mosquito study. In Fig- 1.0 7’ (characters standardized ) Q 5 1.0 15 2.0 2.5 d (characters standardized ) Fic. 11—Two-way frequency distribution of correlation and distance co- efficients for each pair of species in the bee study by Michener and Sokal (1957). Both coefficients of similarity are based on standardized characters. Explanations of this figure and scale of the solid circles as in Figure 4. THE UNIVERSITY SCIENCE BULLETIN "G OINSY Ul SB SapOMO PI[OS oy} JO av[BOS pur oINSY sty} Fo suoneu -E[dxy ‘Stajov1eyo pezipiepurys uO poaseq aie syUoToyyeoo ARIUS IQOg *(Z9BL) F[YOU Aq Apnys oyMbsour yy ur sotoads jo aed yora IOF syUsIOYJood aouLsIp pur UOP{eLI0N Jo UOFNGISIp AoUsNbaif APM-OMT—ZT “OLY ( pezipappuDjs sseyoDsDYO ) p v2 CARA Od 8) 91 ll ol 01 ° @e @e e ° @e@e-> Po ; @@Go0ec-: 28 ©00GO oo 6.6 ° @@Oe@eo- iS ° ° ©°@e@e@eeee-:- = : © © @e@@e@e: : eS -2© e@@@0c0c .- : - £02 @e@e-: g ° e ee ee ~@e0a@: - i a 4 ae fe i 6 | 3 e ° e ee @eree-e e 22 ) e e@eee n e eo 2+ © @ @ se e e e e@ @e @e ee eo «© ee @ ° ° be ° O10 ° ° oO ° ° ° ° e e ° ° 9' NUMERICAL TAXONOMY INVESTIGATIONS 25 ure 12 this can be seen by the presence of a secondary mode on the right side of the two-way frequency distribution. These deviant species seem to differ from the rest more in “size” than in “shape.” Since the majority of the characters used in the mosquito study cannot be interpreted, even indirectly, as a measure of “size,” it is difficult to interpret the meaning of the “size” component in this study. Its presence is indicated by the fact that there is a difference between those relationships indicated by correlation coefficients and distance coefficients. V. SUMMARY OF RECOMMENDATIONS It is recommended that, except in studies based upon characters whose states cannot be linearly ordered, the characters should be standardized by transforming all characters in a given study to a comparable scale. It is suggested that a transformation be used so that each character will have a mean of zero and a standard deviation of unity. It is also recommended that if most of the characters used in a study are measurements of various parts of an organism, and if the OTU’s differ much in over-all size, then corre- lation coefficients should be used as measures of taxonomic simi- larity. If the characters can be converted to ratios independent of size then distance coefficients can be used. If there is a mixture of types of characters then it is difficult to say what the best pro- cedure would be. Until more experience is gained on this point, we would suggest that both coefficients should be computed and comparisons made, since both are valid measures of similarity. 26 THE UNIVERSITY SCIENCE BULLETIN LITERATURE CITED Cain, A. J., and G. A. HARRISON. 1958. An analysis of the taxonomist’s judgment of affinity. Proc. Zool. Soc. Lond., 131:85-98. CaTre.., R. B. 1949. 1, and other coefficients of pattern similarity. Psychometrika, 14: 279-298. Davn, F. N. 1938. Tables of the ordinates and probability integral of the distribution of the correlation coefficient in small samples. Cambridge, Univ. Press, 55 pp. HALTENORTH, T. 1937. Die verwandtschaftliche Stellung der Grosskatzen zueinander. Z. Siugetierkunde, 12:97-240. KENDALL, M., and A. STuART. 1958. The advanced theory of statistics. Vol. 1: Distribution theory. Charles Griffin & Co., London. xii + 433 pp. MicHENER, C. D., and R. R. SOKAL. 1957. A quantitative approach to a problem in classification. Evolution, 11:180-160. PENROSE, L. S. 1952. Distance, size and shape. Ann. Eugenics, 18:337-343. Rocers, D. J., and T. T. TANIMOTO. 1960. A computer program for classifying plants. Science, 132:1115- 1118. Rouir, F. J. 1962. A numerical taxonomic study of the genus Aedes (Diptera: Culici- dae) with emphasis on the congruence of larval and adult classi- fications. Ph.D. Thesis, The Univ. of Kansas, Lawrence, Kansas. 98 pp. 1963. Congruence of larval and adult classifications in Aedes (Diptera: Culicidae). Systematic Zool., 12:97-113. Smirnov, E. ; 1925. The theory of type and the natural system. Z. indukt. Abstamm. Vererbungsl. 37: 28-66. SNEATH, P. H. A. 1957. The application of computers to taxonomy. Jour. Gen. Microbiol. 17:201-226. SNEATH, P. H. A., and R. R. SOKAL. 1962. Numerical taxonomy. Nature 193:855-860. SOKAL, R. R. 1960. Die Grundlagen der numerischen Taxonomie. Proc. 11th Intern. Congr. Ent. 1:7-12. 1961. Distance as a measure of taxonomic similarity. Systematic Zool. 10:70-79. Soka, R. R., and C. D. MICHENER. 1958. A statistical method for evaluating systematic relationships. Univ. Kansas Sci. Bull. 38: 1409-1438. NUMERICAL TAXONOMY INVESTIGATIONS OT SoKAL, R. R., and P. H. A. SNEATH. 1963. The principles of numerical taxonomy. W. H. Freeman and Co., San Francisco. xvi + 359 pp. ZUBIN, J. 1938. A technique for measuring like-mindedness. Jour. Abnorm. Soc. Psychol., 33:508-516. THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vor. XLV] June 7, 1965 [No. 2 The Generic Relationships of the Family Pyemotidae (Acarina: Trombidiformes )' BY EarLe A. Cross 2 ABSTRACT: The tarsonemine family Pyemotidae includes a group of over a hundred and twenty known forms, some of which are of medical or agri- cultural importance. It is thought that only a small fraction of the species are known. The purpose of this study is to add to and synthesize the present knowledge of the history, sytematics, morphology, distribution, habits, and phylogenetic relationships of the group. Classification is largely done at the supraspecific level, species usually being described only as types of the higher categories. The family is divided into 3 subfamilies, one of which (Pygmephorinae ) is further separated into 3 tribes. Twenty-nine genera and subgenera are described as new or redescribed as follows: Dolichocybe, Pavania, Pyemotes, Caraboacarus, Paracarophenax, Acarophenax, Adactylidium, Trochometridium, Siteroptes (Metasiteroptes), S. (Siteroptes), S. (Siteroptoides), Resinacarus, Pediculaster (Pediculaster), P. (Acarothorectes), Pygmephorus, Microdispo- dides, Microdispus (Premicrodispus), M. ( Microdispus), Myrmecodispus, Gly- phidomastax, Perperipes, Peponocara, Parapygmephorus (Sicilipes), P. (Parapygmephorus), P. (Petalomium), Acinogaster, Pseudopygmephorus, Neopygmephorus (Neopygmephorus), N. ( Allopygmephorus). Sixteen new species are described: Paracarophenax dybasi, Adactylidium beeri, S. (Metasiteroptes) macer, S. (Siteroptoides) absidatus, Trochometn- dium tribulatum, P. (Acarothorectes) curculionium, Microdispodides pholi- dotus, M. (Premicrodispus) chandleri, Myrmecodispus dorylinis, Glyphido- mastax rettenmeyeri, Perperipes ornithocephala, Peponocara cathistes, P. (Sicilipes) halictinis, P. (Parapygmephorus) natalensis, P. (Petalomium) krezali, Acinogaster marianae. 1. Contribution No. 1173, Department of Entomology, The University of Kansas. A portion of this work was financed from The University of Kansas General Research Fund, Project No. 4513-5570. 2. Present address: Northwestern State College, Natchitoches, Louisiana. (29) 3—1367 30 THE UNIVERSITY SCIENCE BULLETIN TABLE OF CONTENTS PAGE INTRODUGTIONG) 25.5 S00 Zaes ties, OW. Nhe a Ee ee ee 30 ACKNOWLEDGMENTS): 0c. eee BE A a et ae ate eee ae 34 IECHNIQUES AND PROCEDURES.) -3))¢.1022 40) 000 ee 35 ATS TOR Ys et ss Bes eh Eee eign eae Nc ac eet ee 38 COMPARATIVE, MORPHOLOGY <.)9005 fs. oe oh ote oe ee 40 PEYIEOGEN asco 3 ca hdte: Jen ere Ges a a ee 74 CLASSIFICATION - 32%, .yl0NGS A, mek. fake ale ete ae 84 RITERATURE -@WPED = cece eae i: GS pe Dee 232 INTRODUCTION The purpose of this paper is to add to and synthesize the knowl- edge of the systematics, morphology, habits, and phylogenetic relationships of this little-known group of mites. This paper represents the second major revision of the family, being preceded by the excellent monograph of Krezal in 1959. It should perhaps be pointed out here that the two papers evolved nearly entirely independently and it is therefore a pleasure to note the many points of agreement between them. Every effort has been made herein to correctly credit those ideas originating with the former author. In a real sense, the two works complement one another, Krezal’s stressing work at the species level, this paper being more concerned with higher categories. The family is here divided into 22 genera, these further divided into 29 subgenera and approximately 120 species. This evaluation was made after the inspection of approximately 2400 slide-mounted specimens and on the basis of a careful survey of the literature. In contrast to that of the Tarsonemidae, the classification of the members of this family is. based primarily on the females, a circum- stance resulting from the relative availability of the sexes. In at least the lower forms (Pyemotes, Siteroptes, Acarophenax, Paracaro- phenax) this is due not only to the sex ratio (five to ten percent males ) but also to the transience of the males. Because only a fragment of the total fauna is known and because of the relative scarcity of long series, satisfactory definition of both species and higher categories is sometimes quite difficult. In all cases categories have been defined using as many characters as possible, thereby placing a minimum of reliance upon so-called “weighted” characters. The limited use of such characters has been unavoidable, however, since the author has preferred to erect a phylogenetic as opposed to a “static” (in the sense employed by Sokal in Michener and Sokal, 1957) classification or one based upon “overall basic similarity” (Bigelow, 1958). Data are presently THe Famity PYEMOTIDAE ul being gathered in order to compare such a classification with the one presented herein. Both the size of the morphological gaps between species and the number of species included in a group were considered in erecting taxa. The establishment of monotypic genera was avoided except in cases (e. g., Glyphidomastax) in which the monotypic form showed an unusual amount of morphological isolation. Certain genera, notably in the Neopygmephorini, are difficult to define because of the continuum of characters occurring between the various species. As pointed out by Michener (1944), however, such conditions are to be expected among continually evolving species groups and do not necessarily invalidate the classification. Any specific characteristic, morphological or otherwise, which varied among species, was considered to be useful as a “taxonomic character.” Variation. A great amount of variation in size is present in both sexes, this reflecting itself in various body and setal dimensions. At least the latter may vary independently, e. g., the dorsal setae of female Trochometridium are nearly equal in size although the largest specimen is a third bigger than the smallest. Interpopulational variation in size does not differ markedly in extent from intrapopulational variation insofar as it is possible to judge. The author has often observed phoretic females of pyemo- tids or scutacarids from the same host specimen, therefore pre- sumably from the same population, to differ greatly in this respect. In the case of the scutacarid Imparipes americanus Banks, such variation was seen to occur within the single earthen cells of certain bees and among individuals that may have been the progeny of one mother. In the same species, which is normally heavily pigmented, whitish or very pale individuals occasionally appear, a phenomenon known to occur in other mites, e. g., Dermanyssidae. Setal attributes other than size, such as number and location, remain remarkably stable within a species and are therefore used extensively as characters at specific and at higher levels. The variation of certain characters, such as amount of pigmentation, punctation, development of apodemes, etc., is often not useful be- cause it is inseparable from extrinsic variation due to clearing and mounting. Distribution. The effect of the distribution of pyemotids upon their classification has not yet been evaluated thoroughly. Early acarologists, e. g., Berlese and Jacot, bound by the morphological By THe UNIVERSITY SCIENCE BULLETIN species concept, ignored distribution altogether and considered forms from different continents to be conspecific if they were mor- phologically indistinguishable. Later specialists, aware of criteria of gene exchange in species separation, and certain that oceanic barriers must be limiting, often declared them to be distinct despite supposed inseparability on morphological grounds. Excluding those species undoubtedly spread through commerce (e. g., grain-inhabiting forms of Pyemotes and Acarophenax tribolii), it is evident that many, perhaps most pyemotid genera are widely distributed. Acinogaster, for example, occurs in Angola and Pan- ama, and Neopygmephorus and Premicrodispus are both found in Europe, North America, and Australia. Numerous other examples are available. Widespread distribution of species is seemingly less frequent. Comparison of various North American populations of Pseudopygmephorus sellnicki (Krezal) and the scutacarid Scuta- carus acarorum (Goeze) with the descriptions and statements of variation among European specimens of these forms leads one to conclude that equal or sometimes greater morphological variation may occur in these species between nearby localities than between continents. A similar situation, based upon small sample sizes, seems to exist in certain species of Premicrodispus and Neopygme- phorus. Within North America, certain forms of Pediculaster and Siter- optes appear to vary interpopulationally in an irregular and recur- ring “mosiac” manner rather than in any clinal fashion but the data are too incomplete to permit definite conclusions to be drawn. Although the vagility of mites is largely unknown, particularly over long distances, it seems certain that they are well represented in the air-borne fauna. Lindroth (1957) states that the microfauna is “above average’ in respect to its presence in air samples although he cites no specific data. Gislén (1940) concludes that no geomor- phological barriers exist for microscopic and semimicroscopic ani- mals, their distribution being therefore dependent entirely upon ecological factors. Such species are therefore likely to be cosmo- politan within their ecological limits. It therefore seems not at all reasonable to recognize two species as distinct only because of their occurrence on different continents. My admittedly incomplete data suggest that species in the Pyemo- tidae can easily be explained by such “continuous” distribution. Such a distribution is obviously of little value in determining the relative antiquity of the various groups. THE Famity PYEMOTIDAE 33 Life Histories. The following briefly summarizes some aspects of the known life histories of mites in the family. Separate, more elaborate treatments may be found herein following the generic descriptions. Gravid adult females may (Pyemotinae, Acarophenacinae, some Pygmephorini) become hugely distended and spherical or (most Microdispini, Neopygmephorini) such females may undergo only a slight size increase. Those of the first type usually contain con- siderable numbers (often over 100) of well developed ova. Females of the latter type rarely contain more than three well developed ova simultaneously. In some Pyemotinae (Pyemotes, Trochometridium) and in the Acarophenacinae, the ova hatch into quiescent, indefinite forms which telescope all immature stages and terminate with the appear- ance of the adult mite. This development usually occurs within the distended maternal body so that the adult female gives birth to adults of the next generation. Rarely (Trochometridium) emergence of a new generation of adults occurs after death and partial decomposition of the mother’s body. In Siteroptes, Resina- carus, and at least in most Pygmephorinae, a definitive larval stage is present. Except in the first-named genus, this stage is not known to be active and does not hatch from the egg but develops from larva to adult in the manner described for Pyemotes (above); how- ever, at least some scutacarid females lay eggs which hatch into active larvae, these soon becoming quiescent and moulting directly to adults. The larvae of Siteroptes may hatch and actively move about, sometimes escaping from the body of the mother. Reuter (1900) states (probably erroneously) that a single nymphal stage is present in Siteroptes. Nymphal stages are otherwise unknown in the family. Omission of the various definitive immature stages greatly accelerates embryogenesis and, at least in the Acarophena- cinae, development from egg to adult may consume only three days. Birth of the young adults may occur continuously through the vulva (Pyemotes) or (probably) continuously through a secondary oviporus (Perperipes) or nearly simultaneously in conjunction with the hysterosomal breakdown of the mother (Siteroptes, Acaro- phenax, Paracarophenax ). Females are usually mated immediately after freeing themselves from the enveloping embryonic membranes and often within the body of the mother. In species practicing phoresy, they are the 34 THE UNIVERSITY SCIENCE BULLETIN only known phoretic stage; they probably are responsible for the dispersal of the species in any event. Unmated females of Pyemotes and Siteroptes rarely produce off- spring; if so, such females are arrhenotokous. The number of males is small, rarely exceeding five per cent of the total number of offspring in Pyemotes, Siteroptes, Acarophenax, and Paracarophenax. A higher percentage of males (probably up to 45 per cent) may occur in Trochometridium. Males usually complete their development before females and may then be born (Pyemotes ) or may move about within the body of the mother searching for mates (Acarophenax, Paracarophenax, usually Siteroptes). In the latter case they may initiate or accelerate the hysterosomal break- down resulting in the mass birth of the enclosed mites. Males of Pyemotes remain in the vicinity of the vulva after they are born and copulate with the females as they emerge. Such males may actually assist in the birth of the females by pulling them through the birth opening. They seemingly are parasites upon the body of the mother mite and die quickly if removed therefrom. Food habits of known species are variable although generally constant at the generic level. Siteroptes and Resinacarus are prob- ably fungivorous, Pyemotes is a parasite of various immature insects, Trochometridium is known thus far only as a parasite of larval Apoidea, and Acarophenax and Paracarophenax are egg parasites of beetles. To judge from morphological specializations and from collecting data, it is likely that Glyphidomastax and Perperipes are parasites of immature doryline ants and that most members of the Neopygmephorini are scavengers or fungus feeders. ACKNOWLEDGMENTS The author gratefully acknowledges the advice and criticism of Drs. R. E. Beer, C. D. Michener, and J. H. Camin, all of The Uni- versity of Kansas. Thanks are also due to Drs. E. W. Baker of the U.S. National Museum and G. W. Krantz of Oregon State University for criticism of the keys. The following persons lent specimens from their personal collec- tions or from collections in their care: W. T. Atyeo, University of Nebraska, Lincoln; E. W. Baker, U. S. National Museum, Washing- ton, D. C.; R. E. Beer and R. C. Funk, The University of Kansas, Lawrence; L. Chandler and N. Wilson, Purdue University, Lafay- ette, Indiana; D. A. Crossley, Oak Ridge National Laboratory, Oak Ridge, Tennessee; H. S. Dybas, Chicago Natural History Museum; THE Famity PYEMOTIDAE 35 G. W. Krantz, Oregon State University, Corvallis; G. Lombardini, Florence, Italy; H. B. Morlan, Pensacola, Florida; A. E. Pritchard, University of California, Berkeley; C. W. and M. E. Rettenmeyer, Kansas State University, Manhattan; G. W. Wharton and D. E. Johnston, Ohio State University, Columbus. Dr. H. F. Stammer, University of Erlangen, Erlangen, Germany kindly loaned an unpublished manuscript concerning mites of the family Scutacaridae. This manuscript has since been published ( Karafiat, 1959). Diane A. Hanson of The University of Kansas and David T. Henry of Northwestern State College inked and otherwise ably assisted with the illustrations. Dorthy S. Cross typed the manu- script and Sylvan W. Nelken and Charles H. Wommack of North- western State College otherwise assisted in its preparation. TECHNIQUES AND PROCEDURES Collecting. Many species of pyemotids are found in moss, grass, litter, mammal and bird nests, humus, punky wood, and in other forms of vegetation, living or dead. Such specimens are collected into 70 per cent ethyl alcohol from a berlese funnel. The bodies of various insects are also good sources of material. Mites from these may be removed to alcohol with various moistened implements. Dead specimens from such sources are extremely fragile and should be relaxed before removing. Freshly-killed vertebrates were first chilled then washed in the manner used by Lipovsky, Crossley, and Loomis (in litt.), a method which proved quite successful in the recovery of certain pygme- phorines. A fine camel’s hair brush is useful in the handling of living material. Mounting. Various methods of clearing and mounting were at- tempted. The two which proved most satisfactory were, (1) clear- ing in lactophenol and mounting in polyvinyl alcohol-lactophenol (PVA) as described by Lipovsky (1953), and (2) clearing in glacial acetic acid-chloral hydrate solution and mounting in Hoyer’s me- dium. Providing the mount is not overly thick, the latter medium seemingly possesses the most desirable optical properties and, in addition, allows the preparation of excellent mounts if the specimen is gently heated after the application of the cover slip. It has the disadvantage of crystallizing in a short period of time, particularly in relatively dry climates. 36 THE UNIVERSITY SCIENCE BULLETIN Mites mounted in PVA are relatively more permanent but are less desirable for study than those in Hoyer’s medium and do not seem to expand as well when heated. Viewing. Observations and dissections were made of living speci- mens as well as specimens preserved in 70 per cent ethyl alcohol and in Carnoy’s fixative. The latter proved excellent as a preserva- tive for gross dissections, but dissected parts often proved difficult to clear for observation under high magnification. A similar diff_i- culty was met while attempting to clear certain specimens preserved for long periods of time in ethyl] alcohol. A Leitz binocular microscope having a maximum power of 216 magnifications was used for the above dissections and gross obser- vations. Slide mounts were observed using a phase microscope having a maximum magnification of about 1450 diameters. Most study was done at 970 & and 1450 x. Measurements. Measurements were taken with a .l-mm. grid micrometer placed within a 10 & ocular and were made under the highest magnification possible, 430 or 970 «, depending upon the size of the structure involved. Estimations were made to the nearest tenth of a grid square. Independent measurements of the same specimens, repeated later, suggest that the extrinsic error involved is about + .2 of a square (+ 5y at 430 X, = 2u at 970 ). The mean was obtained for all measurements in cases where n > 5; its standard error was calculated only in those cases where nes LO: The various measurements were as follows: 1. Length of body: Along sagittal line from anterior margin of propo- dosoma to posterior margin of hysterosoma. 2. Width of body: Widest point of hysterosoma measured along trans- verse axis of body. 3. Length of gnathosoma: Along sagittal line from palpal apex (palps in folded position ) to posterolateral condylar processes. 4. Width of gnathosoma: Widest part of gnathosomal capsule measured along a transverse axis. 5. Distance between internal pseudostigmatal sockets: That between cen- ters of origin of pseudostigmatal pedicels. 6. Lengths, legs I to III: Along longitudinal axis of leg from most basal point of trochanter to base of claw(s). 7. Length, leg IV: Along longitudinal axis of leg from most basal point of coxa to base of claws. Because of the curvature of these legs in the more specialized genera this measurement was often divided into two or more sections. 8. Lengths, leg segments: Greatest possible measurement along longi- tudinal axis of segment. Tarsal measurements exclude claw(s) and pulvillus. THE Famity PYEMOTIDAE BL 9. Width, trochanter: Measured along transverse axis at apex of trochanter. 10. Width of coxa IV: Width along transverse axis at medial coxotro- chanteral condyle. 11. Distances between setae are measured as the distances between the centers of the areoli. 12. Distances between apodemes I and II, and between apodemes II and the posterior marginal apodemes, are measured along the anterior median apodeme. Other measurements (e. g., that between the anterior sternocoxal condyles of coxae III) are self-explanatory, or are used only rarely and explained at the place of their appearance in the text. In addition to the errors in measurement due to estimation (see above) certain other sources of extrinsic error should be mentioned. Perhaps the most obvious of these is the distortion of body shape that occurs in preparation of a slide mount. Such distortion is per- haps greatest in width, but is also considerable in length, the more so because of body segmentation in the species concerned. A second source of distortion is due to the position of the mite on the slide, this distortion often imparting an incorrect impression of the shape, size, or location of the structure concerned. Other sources of error may be due to differences in type and amount of clearing and in the kind of mounting medium. Differences in appearance caused by any of the above may overlap natural differences, thus making the detection of intrinsic variation difficult. Of course every effort has been made to take into consideration artificially produced differences and to recognize the naturally occurring differences. Because they serve as a classificatory framework in a poorly- known group, generic descriptions are purposely overdetailed. This detail should be of assistance in pointing out characters of value to future workers. The descriptions of new species are based upon the type speci- mens, variation encountered within the type series being indicated in parentheses. Structures were drawn to scale but setal lengths and thicknesses were estimated and the reader is referred to the descriptions for detail. In a few cases where they could not be included without causing probable misinterpretation, one or more tarsal setae have been omitted from the illustrations of whole mites. Such omissions will not hinder identification of the forms concerned. All setae have been illustrated in the separate, enlarged drawings of tarsi. For those forms in which the types were not available and which 38 THe UNIVERSITY SCIENCE BULLETIN were inadequately described, the author has purposely tended toward conservation in the retention of species names, hoping thereby to keep nomenclatural confusion to a minimum. The reader is referred to the species lists following each generic description for generic assignments of species. THE FAMILY PYEMOTIDAE: HISTORY Although the family was not formally recognized by Berlese until 1898, certain species now placed in the genus Pyemotes were among some of the earliest known acarines. The distinctive, enormously gravid females were noted by Hartig (1834), von Heyden (1835), and DeFilippi (1861) on larval Hyemoptera and by Debey (1849) on larval Coleoptera. Many other authors reported them as para- sites of various insect larvae, sometimes because they proved ex- tremely destructive to cultures of species under investigation (Lich- tenstein, 1868, 1869, 1875). Closely related species were found, at about the same time, to cause the urticarioid dermatitis widely known among workers with stored grains in central Europe and America and commonly re- ferred to in this country as the “grain itch.” Perhaps the first as- sociation of this affliction with mites was made in 1838 in France by P. Debia, who incorrectly assumed it to be caused by a mite belonging to the family Acaridae. A large amount of literature has been and continues to be concerned with the medical importance of members of the genus Pyemotes. As mentioned by Beer (1954), various authors have disagreed on relationships between tarsonemoids and other mite groups. With few exceptions these disagreements have been between those au- thors believing them to be most closely related to the sarcoptiforms and those believing them closer to trombidiform types. The fol- lowing account briefly states the historical development of classifi- cation in each of the two groups. Newport placed his Heteropus (1850) in the family Sarcoptides Latreille. In 1877 Kramer erected the family Tarsonemidae to in- clude the single genus Tarsonemus Canestrini and Fanzago, 1876. This family was placed in his Acarina Tracheata. In 1880, R. Canestrini placed Pygmephorus Kramer, 1877, in his family Oriba- tidi, considering the Tarsonemidae to be a separate sarcoptiform subfamily. This system was followed by Michael (1880) and by Haller (1881 and 1882). Berlese (1885) was the first to recognize the close relationship between Tarsonemus and Pygmephorus, plac- THe Famity PYEMOTIDAE 39 ing them, together with Disparipes (= Scutacarus ) in the subfamily Tarsonemidi of his family Oribatidi. The following year he became aware of Pediculoides Targioni-Tozetti, 1878, and after giving a history of the genus (= Heteropus Newport, 1850, = Sphaerogyna Laboulbéne and Megnin, 1885) added it to the Tarsonemidi. He regarded the subfamily as transitional between his Oribatidi and Sarcoptidi because of the pseudostigmata, the strong ventral apo- demes, and (mistakenly) a ventral genital opening. In 1895, 1897, and 1898, the same author realized the distinctness of the group and placed it in a separate suborder, the Heterostigmata, this sub- order to consist of species in which only the females possessed stigmata. Later in 1898 he separated the Pediculoididae (— Pyemo- tidae Oudemans, 1937) from the Tarsonemidae, the former to in- clude Pediculoides, Pygmephorus, and Podapolipus Rovelli and Grassi, 1888. His 1925 classification is similar except that he has followed Paoli (1911) in removing the genera Disparipes, Diver- sipes, and Imparipes to the family Disparipedidae (= Scutacaridae Oudemans, 1917). The principal supporters of the earlier views of Berlese were Banks (1905a, 1915), and Ewing (1909, 1929, and 1939; vide Beer, 1954). Laboulbene and Megnin (1885b) described Sphaerogyna as a new name for Heteropus Newport, at the same time giving the first resume of the history of the genus. They placed it in the family Trombidiés near the tribe Cheyletides because of the structure of the gnathosoma and palps, the styliform chelicerae, and the posi- tions of the legs. Similar systems were followed by G. Canestrini in 1888 and 1891, and by Trouessart (1896). The latter, on the basis of chelicerae, palps, and (erroneously) the male genital open- ing, placed the group as the section Tarsonemeae of his subfamily Cheyletinae. Brucker (1901) used characters of the chelicerae, integument, and internal anatomy (the latter largely from the midgut and tracheae) as well as habits to erect a classification roughly parallel to that of the two preceding authors. In 1909, Reuter used es- sentially the same characters to erect the subordinal classification in use by most authors today. Within this classification tarsonemines were placed in the single family Tarsonemidae of the superfamily Trombididoidea. Vitzthum (1929) recognized the three tarsonemine families of Berlese (1925), placing them in his supercohort and cohort Tarso- nemini of Reuters Trombidiformes. Oudemans (1931) removed 40 Tue UNIvERSITY SCIENCE BULLETIN Podapolipus and several related genera to form a fourth family, the Podapolipodidae, otherwise retaining the classification of Vitz- thum. Since 1929 no important change in classification has been made, although important papers have been written on the group by Vitzthum (1943), Baker and Wharton (1952), Beer (1954), Cunliffe (1955), and Krezal (1959b). A unique classification, based solely upon stigmatal position, was proposed by Oudemans (1906). In this scheme the family Tar- sonemidae formed a single group, the Trachelostigmata, closely related to both the Antistigmata (including the Uropodidae and Holothyridae) and the Stomatostigmata (including the single family Labidostomidae). This classification has been largely ignored by succeeding authors. Excellent accounts of the history of the generic name Pyemotes were given by Oudemans (1936, 1937a). COMPARATIVE MORPHOLOGY Because of the lack of prior studies within the families of the Tarsonemoidea, the following section is offered, not as a principal area of investigation, but as a guide for more serious students of morphology. The studies described below are limited to members of the family Pyemotidae, but there is little difference between the more special- ized neopygmephorine members of this family and members of the family Scutacaridae and the following observations therefore are pertinent to the consideration of either family. FEMALE Gnathosoma The anteriormost division of the body is the gnathosoma, a cap- sular structure probably derived from both palpal and cheliceral sources. In most pyemotids, as in the other Tarsonemoidea, it is free, protruding from and articulating basally with the propodosoma. In the more specialized genera of the Acarophenacinae, however, it is contained completely within and protrudes only slightly or not at all from the anteroventral portion of the propodosoma (figs. 43-50). In order to simplify the presentation of the following observations, these specialized acarophenacines will be dealt with separately following the discussion of the more ordinary forms. Forms with gnathosoma free. In these forms, the gnathosoma is typically prognathous except in Peponocara and Caraboacarus (figs. THe Famity PYEMOTIDAE 4] 41 and 78) in which genera the lateral and anterior margins of the gnathosomal capsule are expanded ventrally to displace the trophi. In Caraboacarus, this displacement is only slight but it is pro- nounced in Peponocara, P. cathistes appearing distinctly hypog- nathous. In dorsoventral aspect, the gnathosomal capsule (figs. 1, 2) is a variably shaped (globose, quadrate, rectangular, or ovoid) struc- ture, its basal margin well sclerotized and produced lateroventrally into a pair of condylar processes. Probably only dorsoventral movement of the gnathosoma upon the propodosoma is possible. A dorsal, longitudinal, median apodeme may join the posterior margin, extending anteriorly for varying distances. The dorsal and ventral surfaces of the capsule extend anteriorly between the palpi and enclose the mouth parts from above and beneath. Dorsally, the gnathosomal capsule usually bears two pairs of setae, the internal and external dorsals (fig. 1), of which the internals are always the most anterior. In Pyemotes and in Dolichocybe these setae arise well forward, the internals well in front of or in line with the second palpals. In all other genera the internals are posterior to the second palpals. In Caraboacarus and in certain species of the Pygmephorinae the external dorsals appear on the ventral surface. In Microdispus (Premicrodispus) the internal dorsals may be absent, and in at least one species of Pygmephorus both pairs of dorsals are lacking. Arising laterally or dorsolaterally behind the palpi there may be one or two pairs of postpalpal setae. These usually lack distinct areoli and may be hidden along the lateral gnathosomal margin. They are especially difficult to see in certain Pseudopygmephorus and may be lacking in some Premicrodispus. Ventrally, a single pair of gnathosomal ventrals (fig. 2) usually arises from the anterior half of the gnathosoma. Trochometridium and Caraboacarus have a second pair of gnathosomal ventrals, those of the latter genus being heavily sclerotized and spheroid in shape. In most prognathous forms (exception, Pavania) the palpi arise from lateral excavations of the gnathosomal capsule, their lateral margins forming a continuous line with the lateral margins of the capsule. They enclose the trophi at the sides. Basally they articu- late with the capsule, being capable of at least lateral-mesal move- ment in all genera except Pyemotes and Caraboacarus, in which they are inarticulate and at least partly fused to the wall of the gnatho- soma (Krezal, 1959, states that the palpi of Pyemotes are distinct 49, THe UNIVERSITY SCIENCE BULLETIN and “slightly movable” but I am unable to discern either free mar- gins or articulations in the specimens available to me). In these genera the line of fusion is more or less distinct and both pairs of palpal setae are retained. In Pavania, the palps are free and arise from the ventral surface of the capsule. In forms having free palpi, these appendages are usually at least slightly palmate and possess one or more apical teeth. The teeth usually arise from the medioventral surface of the palp and may be free and seemingly movable (e. g., Pygmephorus) or fused to the palpal apex. In Dolichocybe and in some Neopygmephorus (fig. 2) the palpal teeth may be replaced by a clawlike hook. They are reduced in Dolichocybe and Trochometridium and absent in Pyemotes and Caraboacarus. In Dolichocybe, the palpi are sub- cylindrical, while in Glyphidomastax, Perperipes, and some Myrme- codispus they are greatly elongate and sharp. In all forms seen by this author the palpi are one-segmented although in some forms (fig. 2) suturelike lines may occur on one or more surfaces. Oudemans (1931) mentions, probably mis- takenly, forms with two palpal segments. The palps of Pygme- phorus thienemanni Willmann (not seen by this author) are said to possess four segments; this form is seemingly highly specialized in other respects. Universally present in all forms having a free gnathosoma, and arising from the dorsolateral, lateral, or ventrolateral margins of the palpi, are the conspicuous first and second palpal setae, the former the most basal. These often appear to divide the palps into three subequal areas, and, because of their constancy in location, may provide information concerning areas of differential growth. In addition, the inner apical surface of the palp may possess one or two tiny setae, these not always visible. Arising from the ventral surface of the apical half in many forms are palpal solenidia 1 and 2. The latter is atypical in appearance, being non-striate. It may be cylindrical, clavate, capitate, cuspidate, or concave api- cally, and is nearly always much larger than solenidium 1. The latter is typically striate and arises lateral or posterolateral to 2. In Pavania, Pyemotes, and Trochometridium, at least solenidium 2 is reduced and may be indistinct. In Caraboacarus, Perperipes, Glyphidomastax, and some Myrmecodispus, the palpal solenidia may be lacking. Palpal setation frequently offers specific characters, particularly as regards shape and placement of the various setae. THe Famity PYEMOTIDAE 43 The palps probably function in most forms to anchor the mite to the substrate while it feeds. The trophi were not studied intensively for purposes of this work and the following brief account is taken largely from the excellent work of Krezal (1959b) and refers to Pyemotes. The chelicerae are dorsal, their distal portions styletlike, their bases guttate, recurved laterally, housed in anterodorsal pockets which may appear broadly tuberculate upon the dorsal surface of the capsule. Their guttate bases are prominent landmarks of the gnathosoma, often visible when the sylets are indistinct. At least in P. scolyti, the chelicerae are separated by a longitudinal dorso- ventral septum and bounded dorsally and ventrally by chitinized plates. The styli are independently protracted or retracted by muscles originating basally (protractors) or laterally (retractors) from the capsular wall and inserting basally (protractors) and distally (retractors) upon the cheliceral base (“Stechborstenhebel” of Krezal). Beneath the cheliceral opening lies the oral orifice, enclosed apically in an acuminate cone. This cone is inserted into the wound made by the chelicerae once the latter are withdrawn, and the sucking operation commences. The oral cone is contiguous with an enlarged, ovoid, ventral esophageal-pharyngeal pump (“pharynx” of authors). Brucker (1901), in describing the operation of this pump, states that it expands slowly, then contracts quickly to push the liquid food into the gut. Contractions were noted to be on the order of about 50 per minute. With certain modifications, the foregoing is probably similar to the structure and function of the mouthparts of most Pyemotidae. The chelicerae are enlarged and falcate in Pavania, often bladelike in Pyemotes, Trochome- tridium, Pepoponocara, and in all genera of the Acarophenacinae, but are smaller and usually indistinct in all remaining genera. It should perhaps be mentioned that with the exception of Pyemotes, all those forms with enlarged chelicerae are phoretic upon the adults of (and probably parasites of the immature stages of ) various insects. The seemingly rare occurrence of phoresy in species of Pyemotes correlates well with the broad host tolerance seemingly characteristic of that genus. The ovoid esophageal-pharyngeal pump is reduced in Doli- chocybe and absent in all other genera possessing a free gnathosoma. In these latter the flexible esophagus is usually visible posterior to the chitinized buccal chamber. Within the gnathosoma, it may be 44 THE UNIVERSITY SCIENCE BULLETIN straight or characteristically looped, coiled, or spiralled. In the Pygmephorinae, a pair of small, globular or elongate structures are often found adjacent to the esophagus. They appear to connect to it and pernaps are salivary glands. In the Pygmephorini and Neopygmephorini a narrow “finned structure” in the shape of a bow-tie may be visible. If present this structure is always followed by one or two similar structures located in the propodosoma. These will be mentioned in more detail later. The supra- (“Chitinspange’ ) and sub-cheliceral plates (“Chitinplatte”) of Krezal assume vary- ing shapes or are absent in genera other than Pyemotes. Gnathosoma hidden. In Acarophenax, Paracarophenax, and Adac- tylidium (figs. 43-50) the gnathosoma is much reduced and nearly or completely enclosed within the propodosoma, with which struc- ture it may be partially fused. Its structure is therefore difficult to ascertain. The palps of such species are not or only partly dis- tinct and setae of the gnathosomal region are lacking except in Acarophenax, which possesses one to three pairs. The most promi- nent gnathosomal structures in these forms are the enlarged, antero- ventrally or ventrally directed chelicerae and the esophageal pharyngeal pump, both of which are similar to those described above for Pyemotes. In addition to the above, species of Paracarophenax exhibit a unique, elliptical, median ventral structure which may be the paired salivary glands and their heavily sclerotized ducts (the lateral “sclerotized margins” of this structure are indeed tubular). Paired ducts presumed to be salivary are distinct beneath the respiratory system in Acarophenax and Adactylidium. In the region of the dorsal U-shaped apodeme these fuse, the resulting duct appearing to be suspended from a thin median longitudinal apodeme which may be contiguous posteriorly with the U-shaped apodeme. An- teriorly, this duct bends ventrad to enter the anterior portion of the pharyngeal pump at which juncture it may become campanuliform or otherwise dilated. Propodosoma The propodosoma (figs. 3 and 4) is the second body division. Located between the gnathosoma and the hysterosoma, it is the section of the body bearing the first two pairs of legs. Except for the genera Acarophenax, Paracarophenax, and Adactylidium, which will be discussed later, the propodosoma consists of a single struc- ture which, in shape, may be likened to a sleeveless sweater with four instead of two armholes. In the Pyemotinae, its dorsum is THE Famity PYEMOTIDAE 45 about as long as the venter, while in the majority of the Pygme- phorinae the venter is distinctly longer than the dorsum. The anterior circumgnathosomal foramen varies greatly in shape and position, its characteristics greatly dependent upon the manner in which the gnathosoma is normally carried. In those genera (Pye- motes, Dolichocybe, Caraboacarus) in which it is directed straight forward, or nearly so, the foramen opens approximately at a right angle to the longitudinal axis of the body. In some Siteroptes, Resinacarus, and in most of the Pygemophorinae, the gnathosoma is carried at least slightly downward and is often bent directly ventrad or posteroventrad. In these forms the propodosomal dorsum is elongated to cover the gnathosomal base, thus displacing the foramen ventrad. The ventral margin of the foramen may be entire or deeply emarginate to allow for the reception of the gnathosoma in those species in which it is normally carried postero- ventrally. In Pygmephorus and in some Parapygmephorus the lateral margins of the circumgnathosomal foramen may be heavily sclerotized, appearing to form a “collar” for the gnathosoma. In such cases the later is carried nearly vertically with only its dorsum and apex exposed. Whether entire or emarginate, the ventral margins of the circumgnathosomal foramen are always greatly thickened to form apodemes I (Epimerites I of Krezal) which func- tion to provide strengthening for the anteroventral sternocoxal con- dyles of coxae I as well as to provide reinforcement for the fora- minal margin and points of gnathosomal articulation. The entire ventral surface of the propodosoma (fig. 3) is the anterior ventral plate. It is bounded anteriorly by apodemes I and its posterior limits may (Pygmephorinae) or may not (most Pye- motinae) be well defined by the posterior marginal apodemes (Epimera II of Krezal) which are usually most distinct laterally. At the sides, the posterior marginal apodemes bend first dorsally, then anteriorly to provide dorsal support for legs I. In some species of Myrmecodispus and Acinogaster, these dorsal arms may become enlarged and scapuloid. Directed longitudinally and joining apodemes I with the posterior marginal apodemes is the anterior median apodeme (Sternum I of Krezal). This apodeme may be weak, interrupted, or intermittent (various species), but it is invariably present. Apodemes II (Epimera I of Krezal) arise laterally near the anterior sternocoxal condyles of legs II and terminate mesally in or near the anterior median apodeme. In the Pyemotinae, Acaro- 46 THe UNIVERSITY SCIENCE BULLETIN phenacinae, and Pygmephorini they are usually straight or sigmoid and join anterior sternocoxal condyles II laterally. A lateral fold or apodeme, which is continuous with apodeme II, is usually present between the anterior sternocoxal condyle of coxa II and the posterior portion of coxa I. In the Microdispini and Neopygmephorini apodemes II are usually distinctly arcuate. In these tribes, they often terminate mesal to sternocoxal condyles II and integumental folds or thickenings continuous with them often reach the areoli of external ventrals I. Apodemes II are weak or absent in Dolicho- cybe, some Microdispini, and some Parapygmephorus. In a few Microdispini, most Neopygmephorini, and most Scuta- caridae, a secondary transverse apodeme may be present. This apodeme is straight or slightly procurved. It passes near or through the intersection of apodemes II with the anterior median apodeme at about a right angle to the latter. The intersection of these two sutures at the anterior median apodeme delimits between them, on each side, a lateral trianguloid sclerite which bears ventral setae II. In forms possessing this sclerite, external and internal ventrals II are always close together. This arrangement differs from that found in more primitive forms in which the setae tend to be more evenly spaced across the ventrite. The nomenclature of basal acarine leg segments has long been a subject of disagreement among various workers. Vitzthum (1943) believed the femur and genu (or, in females of Tarsonemidae, genu and tibia) to be fused, the definitive leg segments then being coxa, trochanter, femur (or femurogenu ), tibia (or genuotibia), and tarsus. Oudemans thought that the coxae had fused with the sternum, the first definitive segment therefore being the trochanter. Beer (1954) deemed the evidence insufficient to support the view of Oudemans and designated the first definitive segment as the coxa. Krezal (1959) follows Oudemans in believing that the coxae have become flattened and platelike, fusing with the primitive sternum to be- come the definitive ventral plate. This belief is based largely if not entirely on the fact that the ventral apodemes are of such disposi- tion that they may be construed as being lines of fusion between coxal plates. Other points considered in favor of this thesis are (1) musculature and (2) that in certain prostigmate groups the coxa is generally regarded as being fused with the sternum (e. g., Trombiculidae, vide Wharton and Fuller, 1952, Hydrachnellae, vide Mitchell, 1955, and others ). THE Famity PYEMOTIDAE 47 The present author does not deny the possibility of correctness of the foregoing hypothesis but wishes to point out the following: (1) The ventral apodemes undoubtedly do not represent the true lines of fusion between the coxae, as a comparative study of the various genera will show, but vary greatly in shape and extent, at least in part according to the requirements for reinforcement and areas for muscle attachment in a particular form, this in turn de- pendent upon the amount and type of locomotion, carriage of the legs, size and shape of the mite (compare apodemes IV of Peri- peripes and various scutacarids ), and other factors. For example, apodemes IV in Siteroptes, the species of which have functional hind legs, terminate laterally at anterior sternocoxal condyle IV, a point at which strengthening is needed. In the Neopygmephorini and Microdispini, groups in which legs IV are modified and there- fore used sparingly if at all in normal locomotion, the same apodeme terminates in coxal foramen III, usually in its anterior half. More- over, ventral apodemes are the most poorly developed in what seem to be the most primitive genera (Dolichocybe, Pyemotes, Trocho- metridium, Siteroptes) and in the known larvae, although larvae of all genera studied retain such primitive characters as two claws on tarsus I and five-segmented front legs. (2) Although the coxae of certain prostigmate mites appear to be fused with the venter, no evidence of the relationships of these forms to the Tarsonemoidea has been presented. Indeed, specific examples of such forms are not cited by Krezal in his work. (3) It seems to this author premature at this time to make claims based upon myology since no detailed studies have been made for any group in the Tarsonemoidea. Al- though it is true that most muscles responsible for leg movement do arise from the ventral apodemes, this does not indisputably establish coxal origins of the latter. Specimens of Trochometridium tribulatum cleared in xylene and mounted in balsam by this author resembled certain unionicolids investigated by Mitchell (1955), in that the origins of the muscles inserted upon the first definitive leg segment were diverse. Thus, muscles appearing to be promotors and levators of legs I and II of Trochometridium originated side by side from the posterior marginal apodeme and muscles appearing to function as remotors and levators of the same legs originated side by side from the anterior median apodeme well cephalad of apodemes II. In consideration of the above, it is felt that the assumption of a coxal origin of the ventral plates is based upon meager evidence and 48 THE UNIVERSITY SCIENCE BULLETIN should not be accepted uncritically. In any event, the terms (sternum, epimeron, epimerite) proposed for ventral plates and sutures by earlier authors are morphologically incorrect and there seems to be little reason for retaining them other than historical precedent. This author therefore prefers to use, with modification, the nomenclature used by Beer (Table 1). The four sclerites of the anterior ventral plate called coxae by Oudemans and his followers are accordingly designated as ven- trites, those anterior to apodemes II as ventrites I, those posterior as ventrites IT. The anterior ventral plate bears from two to six pairs of setae (fig. 3). If two, four, or six pairs are present, they are equally divided between ventrites I and II. If three or five pairs are pres- ent, either ventrite may possess the odd pair. All members of the Acarophenacinae bear fewer than four pairs. Members of the genus Pyemotes and all members of the tribes Microdispini and Neopyg- mephorini bear four pairs, and the remaining genera bear five or six pairs. These setae are named in the manner designated by Krezal (1959b), except that the author prefers to substitute, for reasons explained above, the word “ventral” for “coxal” (Table 1). Which of the ventral setae drop out or are added in the various genera is not known and the author arbitrarily designates the pair normally present in some Pyemotinae and Pygmephorini but lack- ing in the Microdispini and Neopygmephorini as the median ventral pair. External ventrals I are furcate in some species of Siteroptes, Pseudopygmephorus, Neopygmephorus, and Parapygmephorus, sim- ple in the remaining forms. The propodosomal dorsum is irregularly trapezoidal, elongate, or inversely T-shaped. Its posterior margin may meet or overlap the first hysterosomal tergum (Acarophenacinae, most Pyemotinae) or be overlapped by the latter (Pygmephorinae). In most members of the last-named subfamily it is delimited posteriorly by a dorsal transverse apodeme the ends of which fuse with the dorsal margins of coxal foramina IT. The two variously-shaped peritremes (fig. 4) enclose and are usually distinct from the stigmatal openings. Both are lacking or at least not discernible in Pavania and Dolichocybe, but present in all other genera. The shape, proximity, and position of the peri- tremes are useful in classification at both the generic and specific levels. THe Famity PYEMOTIDAE 49 Posterior to the peritremes arise the pseudostigmata. These two capitate, bladderlike organs are of uncertain function but are almost certainly sensory receptors. Each originates deep within the propo- dosoma from the internal pseudostigmatal socket and usually (ex- ceptions, Pavania and Dolichocybe) passes through the center of a series of concentric ringed structures, the external pseudostigmatal socket, in the region of the body wall. The structure of the rings of the external socket indicates that it is probably capable of re- sponse to the slightest movement of the pseudostigmata. From one to three pairs of dorsal propodosomal setae are present (fig. 4), these often greatly reduced in forms in which the propo- dosoma is covered by the first hysterosomal tergum. The most anterior pair of these setae (“verticals” of Krezal) are nearly always closely associated with the peritremes and are called the stigmatals herein. The most posterior pair (“external scapulars” of Krezal), usually arising mesad or posteromesad of the pseudostigmata and always distinctly longer than the others, is called the posterior pseudostigmatals. The remaining pair usually arises anterior or anteromesal to the pseudostigmata and is designated the anterior pseudostigmatal pair (“internal scapulars” of Krezal). A predicta- ble sequence of loss usually occurs among these setae, the stigmatals nearly always being lost first, followed by the anterior pseudo- stigmatals. The number of pairs of dorsal propodosomals is of value in the higher classification of the family while their size and placement is useful in species determination. In the genera Acarophenax, Paracarophenax, and Adactylidium a unique division of the propodosomal dorsum occurs. In these genera, the superficial dorsal surface is enlarged to form the dorsal propodosomal plate (figs. 46 and 49), its anterior margin covering the gnathosomal region, its posterior margin covering the first hysterosomal tergum. Because of its size and structure, this dorsal plate greatly resembles the enlarged first hysterosomal tergum (“clypeus” of authors) found in the Scutacaridae and in some spe- cialized pyemotids and may easily be confused with it. This parallelism may have led Vitzthum (1943), Hughes (1948), and others to incorrectly assign Acarophenax to the Scutacaridae. A smaller, more internal portion of the propodosoma, the propodo- somal arcula, which is boxlike or nearly so, apparently functions to enclose and, at least in Acarophenax and Adactylidium, seems to fuse with the reduced gnathosoma, providing areas for muscle at- 50 THE UNIVERSITY SCIENCE BULLETIN tachments and points of articulation. Im addition, it bears the peritremes. At least in Acarophenax and Adactylidium this arcula may be partly or largely sclerotized. In these genera it is visible from above as a heavy U-shaped or tridentoid apodeme, the lateral arms of which bear the anteriorly-directed peritremes at their apices (fig. 5). Other dorsal apodemes may arise from the vicinity of the stigmatal openings, proceeding mesally apparently to fuse with the walls of the buccal cavity. The ventral or anteroventral portion of the arcula, which encloses the more or less reduced mouthparts, may be an anterior extension of the anterior ventral plate, in which case the most posterior setae of the gnathosomal region may be the first internal ventrals. In all three of the above genera the arcula appears to enclose the anterior portion of the gnathosomal region, but this point is not clear. The bipartition of the propodosoma apparently allows the dorsal plate to separate from the remaining gnathosomal-propodosomal complex, thus allowing more space for the developing young. Rack (1959) illustrates this condition (pp. 416-417) in Paracarophenax dermestidarum but does not mention it in the text. Pseudostigmata are lacking in Paracarophenax, Acarophenax, and Adactylidium. The dorsal plate bears two or three pairs of setae, while the arcula bears zero to one pair of dorsal or anterior setae and one to two pairs of ventrals. Homologies of the propodosomal setae in these genera are not clear. Internal anatomy. Structures associated with two major systems, the respiratory and digestive, are usually visible in the region of the propodosoma and may be useful as taxonomic characters. The peritremes enclose the stigmatal openings, which are usually circular (most Neopygmephorini) but may be slitlike (Trocho- metridium). In some genera (e. g., Microdispus) the stigmata open anteriorly, their openings being therefore hidden, but in most cases they turn dorsally into the peritremes and their tubular nature is readily observable. These openings are the terminations of the main tracheal trunks which proceed posteromesally into the propo- dosoma, each connecting there with a bulbose internal atrium designated by Krezal (1959b) as division III of the tracheal trunk. This division is strongly sclerotized in some Acarophenacinae, a fact which led Hughes (1948) and others to characterize Acarophenax as possessing two pairs of stigmatal openings. The posterior end of this division appears sievelike, its external surface giving rise to THE FAMILY PYEMOTIDAE 51 TaBLE I.—Terminologies of Beer (1954), Krezal (1959) and Cross for Structures of the Propodosoma Beer, 1954 Krezal, 1959 Cross apodeme I transverse apodeme anterior median apodeme stigmatal openings stigmatal openings epimerite I epimeron IT sternum I anterior sternal plate epimeron I coxae I[-II externae | ‘IT &Il setae coxales mediae internae stigmata stigmata setae verticales setae scapulares internae setae scapulares externae circumgnathosoma! foramen apodeme I posterior marginal apodeme anterior median apodeme anterior ventral plate apodeme II secondary transverse apodeme ventrites I-II external median internal ventrals I & IT dorsal transverse apodeme peritremes stigmatal openings internal pseudostigmatal socket external pseudostigmatal socket stigmata] setae anterior pseudostigmatal setae posterior pseudostigmatal setae dorsal propodosomal plate (some Acarophenacinae) propodosomal arcula (some Acarophenacinae) 52 THE UNIVERSITY SCIENCE BULLETIN a number of fine tracheae. These tracheae may be long or short, probably depending (at least to some extent) upon the species. In Pyemotes they may reach the posterior extremity of the opisthosoma (Krezal, 1959b), in S. (Siteroptoides) they are visible as far as the posterior margin of coxa IV, and in most Acarophenacinae they appear still shorter. Air was seen to issue only from the structures designated herein as peritremes when specimens of Pygmephorus and Neopygme- phorus were drowned in alcohol and probably these comprise the only external respiratory openings. The esophagus is often visible within the propodosoma. In some Siteroptes, and most Pygmephorinae, one to three separate circum- esophageal “finned structures” are distinct within the propodosoma and the anterior portion of the hysterosoma. The anatomy and function of these structures is uncertain at present but Krezal’s conjecture that they are esophageal pumps seems reasonable. The shape and number of these structures and the structure of the fins are sometimes used as diagnostic characters (see generic descrip- tions ). Hysterosoma The largest and most posterior of the three body regions is the hysterosoma. In dorsal aspect, this region usually consists of four distinct, telescoping body segments, the most posterior usually com- pletely overlapping a fifth® segment which may be cuplike or reduced to a ventral plate, or vestigial. Ventrally, the large pos- terior ventral plate bears the third and fourth pairs of legs. Its posterior margin overlaps the platelike opisthosomal venter which, in turn, usually overlaps the ventral portion of the fifth segment. The hysterosoma is movable against the propodosoma, a distinct conjunctiva separating the two. This conjunctiva may arise from the adjacent margins of the two regions or, in those genera in which the first hysterosomal tergum overlaps the propodosoma, the dorsal conjunctiva of the former sclerite may arise from an arcuate trans- verse apodeme behind its anterior margin. In genera in which physogastry involving the entire hysterosoma is pronounced, the conjunctiva between propodosoma and hysterosoma may be visible as distinct pleated bands which unfold as physogastry progresses. 8. Although Krezal states that the hysterosoma consists of only three dorsal segments and a “Chitinkappe,”’ the present author finds (at least in Perperipes, Neopygmephorus, and Pygmephorus) four distinct dorsal segments plus the sclerite designated herein as segment V. The latter may or may not be caplike. He is of the opinion that five seg- ments probably also exist in most remaining genera, although in the more specialized forms of the Acarophenacinae the fifth segment may be absent or reduced to a ventral plate which no longer bears setae. Oudemans (1931) mentions species of pyemotids with seven dorsal segments but these observations are probably incorrect. THE Famity PYEMOTIDAE 53 Dorsum. Sclerites. The first tergum may meet, be overlapped by, (Pyemotinae, Acarophenacinae) or overlap (Pygmephorinae) the posterior margin of the propodosoma. The degree to which the last-named condition occurs varies considerably within the Pygme- phorinae. In some forms, only the posterior margin is covered, while in others the entire propodosomal dorsum is hidden. This latter condition has arisen several times within the Pygmephorinae (Microdispini, some Petalomium, Allopygmephorus) as well as being characteristic of the entire family Scutacaridae. In the latter family the first tergum (“clypeus” of Berlese and others) is espe- cially large, being expanded both laterally and anteriorly to com- pletely cover the propodosoma as well as the basal portions of the gnathosoma and legs I and I. In the Scutacaridae the “clypeal” margins are always free and radially striate (the striae seemingly conjunctival in origin), a condition only recently discovered in two new genera (Perperipes and Glyphidomastax) of Pyemotidae. In all other pyemotid genera, if the first tergum is enlarged, its anterior and lateral margins are neither free nor striate. Tergum I of Trochometridium differs from that of all other genera in that it is not entire, but tripartite, consisting of a single quadrate mesal plate and a pair of lateral plates which overlap the former mesally. Terga II to IV are bandlike and, except for their posterior margins, which may be procurved in some species (e. g., Psewdopygmephorus sellnicki), and for their setation (which will be discussed later ), offer no unusual modification. In a few forms, a pair of rounded, pale areas are visible on segments II or III. These mark the dorsal insertions of the dorsoventral muscles. In some species, a wide band of conjunctival pleats is visible between terga III and IV. Segment V is usually completely overlapped dorsally and laterally by segment IV. In lateral mounts and in dissected specimens it is often seen to be a cuplike cap which encloses the posterior end of the mite. It bears the slitlike apical or apicoventral vulva. In Dolichocybe and Pavania the fourth segment forms the terminal cap. In these two genera its anterior margin is contiguous ventrally with segment V, the latter plate-like and bearing a single pair of short setae which flank the vulva. The anterior margin of segment V is contiguous in turn with a separate, flattened sclerite, the antevulvar plate (fig. 17), unique to these two genera. The latter bears an internal genital sclerite and, externally, two pairs of tiny setae. Its origin is unknown. In Pyemotes and in the more spe- cialized Acarophenacinae, a reduction of the fifth segment is also 54 THe UNIVERSITY SCIENCE BULLETIN found, this accompanied by a reduction or loss of the setae there- from. Setation of dorsum. The setation of the hysterosomal dorsum is generally consistent and perhaps homologous throughout the Tarso- nemoidea and is quite similar to that found in many families of the Prostigmata. Without exception, the first hysterosomal tergum bears two pairs of setae, the second a single pair (fig. 4). The third tergum bears two pairs of setae in all forms except several species of Siteroptes, in which only a single pair may be present. Tergum four bears two pairs of setae except in Metasiteroptes macer, Setation of the fifth segment is variable. One pair is pres- ent in Dolichocybe, Pavania, and Pyemotes, all remaining pyemotine genera possessing three pairs. Acarophenacine genera resemble Pyemotes in having one pair except for Adactylidium and some Acarophenax, in which segment V is without setae. Three pairs of setae are present in all Pygmephorini, in a few Microdispini, and in most (exception, Allopygmephorus) Neopygmephorini. In most Microdispini and in Allopygmephorus only two pairs are present. The number, shape, and position of the dorsal setae are used to separate both species and genera. The author believes the setae of the first four terga to be homologous, and in order to show this homology, as well as for purposes of simplification, further modifies Krezal’s modification of Paoli’ (1911) terminology (Table 11). Venter. Sclerites. Ventrally (except in Pavania and Dolicho- cybe) the hysterosoma consists of but two easily recognizable plates, the posterior ventral plate and the large opisthosomal venter (fig. 3). The former bears the two posterior pairs of legs. If it is prim- itively composed of the sterna of two segments (and in addition, perhaps, the fused coxae), then no line of fusion between these segments exists in the genera presently known. In two forms, Siteroptes (Metasiteroptes) and Pediculaster (Acarothorectes), a transverse division of the posterior ventral plate occurs, but this is undoubtedly a secondary development. In the genera Pyemotes (fig. 21) and Trochometridium (fig. 30) this plate is tripartite, con- sisting of two lateral portions and a mesal portion. The former meet and overlap only in the anterior portion of the plate. Dorsal to these two lateral plates and between their mesal margins is the triangular mesal portion of the plate, easily recognizable by its shape and position and because it bears a single pair of poststernal setae. In all remaining genera the posterior ventral plate is entire. THE Famity PYEMOTIDAE 55) The anterior margin of the posterior ventral plate is entire and recurved. Laterally, each margin consists of a single flap which extends dorsally to cover the lateral margin of the second hystero- somal tergum, but lies beneath that of the first tergum. At least in Acarophenax tribolii each lateral flap is split (fig. 43) along a line terminating ventrally near the anterior margin of coxa II. The posterior portion of this flap covers the lateral portion of the second tergum as described above, while the anterior portion is enclosed by the first tergum. In all other genera the lateral flaps are seem- ingly entire. In those forms in which the posterior ventral plate is entire, its posterior margin may be variously excavated, linear, or lobate. In Dolichocybe, Paracarophenax, some Pseudopygmephorus, and in most of the Microdispini, this margin is tripartite. The posterior ventral plate has four apodemes whose develop- ment and shape are useful in classification. The posterior median apodeme may be present in varying degrees or absent. It corre- sponds to and often appears to be a continuation of the anterior median apodeme. Apodemes III and IV are diagonal or transverse and may be complete or incomplete in the various genera. In pre- sumably primitive Pyemotinae (e. g., Dolichocybe, Pavania, Pye- motes), and in some Acarophenacinae, these apodemes are well developed only laterally and serve to strengthen the anterior sterno- coxal condyles of their respective legs (figs. 13, 21, 43, 45, 48). The posterior median apodeme is usually completely absent in these forms. Assuming the above arrangement to correspond to the primitive condition, figure 7 illustrates a proposed sequence of stages in the evolution of these apodemes. In general, there is a tendency toward the increased development of apodemes III and IV as well as the posterior median apodeme in the higher Pyemotinae and lower Pygmephorinae [some Siteroptes s. s., S. (Siteroptoides), Pediculaster], followed by a decrease in size and distinctness in the higher Pygmephorinae. There is, however, a continuous trend toward the anterior migration of apodemes III and IV in all genera, culminating in the arrangement depicted in Neopygmephorus (fig. 7h). In this genus, apodemes III are entirely or almost entirely lost and apodemes IV have lost their original contact with the anterior sternocoxal condyles of legs IV, their lateral terminations joining the anterior halves of the inner margins of coxae III instead. Apodemes V are present only in Trochometridium and some species ~ 56 THe UNIveRsIry SCIENCE BULLETIN of the genera Siteroptes, Pediculaster, and Pygmephorus. They ex- tend obliquely from coxal foramina IV to the posterior termination of the posterior median apodeme. Probably the loss of the strength- ening apodemes of legs IV and their migration anteriorly is corre- lated with changes with the modification of legs III and IV (p. 64). In all genera but Dolichocybe and Pavania, the opisthosomal venter is a single, large plate whose anterior margin arises beneath the posterior margin of the posterior ventral plate and whose pos- terior margin overlaps the ventral portion of the fifth segment. The shape of the margin of the latter varies intraspecifically in at least some forms. The lateral margins of the opisthosomal venter re- semble those of the posterior ventral plate in being flaplike. They probably cover the lateral margins of the third tergum and lie beneath the second tergum in the manner shown by Krezal (1959b, p. 388) although this cannot be verified in all genera. A band of pleated conjunctiva may connect the posterior ventral plate with the opisthosomal venter. In Pyemotes and in some Acarophenacinae a second band may be present between the opisthosomal venter and segment V. In Dolichocybe and Pavania (figs. 13, 17), the opisthosomal venter is divided into a pair of earlike lateral lobes. These lobes do not meet medially and are separated posteriorly from the antevulvar plate by a distinct, pleated conjunctival band. Their lateral margins were not seen by this author. Setation of the venter. The setal nomenclature of the posterior ventral plate proposed by Paoli and retained by Krezal (1959b) is utilized here without change (Table II and fig. 3). Six pairs of setae occur on this plate in all forms except the following, in which this number is reduced: all of the Acarophenacinae except Cara- boacarus, Pyemotes, certain species of Siteroptes, and certain species of Pediculaster. In the first group mentioned, one to three pairs may be lacking, while the remaining genera lack one or (rarely) two pairs. The internal presternals are always present, arising near or in front of apodemes III (or their remnants) usually near the posterior median apodeme. The first axillaries arise near to well anterior of apodemes III and near the margin of the third coxa. They are invariably present. The external presternals always arise at least slightly behind the internal presternals and distinctly mesal to the first axillaries. In most Pygmephorinae these setae arise near or anterior to apodemes IV. They are rarely absent (some Acaro- phenacinae). The second axillaries arise anteromesal to but near TasLe IJ.—Comparative Terminology of Beer (1954), Krezal (1959), and THe FAamiIty PYEMOTIDAE Cross for External Structures of the Hysterosoma Beer, 1954 Krezal, 1959 Cross apodeme III apodeme IV posterior median apodeme tergites I-III chitinkappe chitinkappe ventral shield I ventral shield II epimerite ITI epimeron III epimeron IV sternum II axillares I axillares II setae praesternales externae and internae setae poststernales externae and internae humerales internae humerales externae dorsales lumbales internae lumbales externae sacrales internae sacrales externae caudales internae caudales externae I caudales externae II terga I-III tergum IV segment V posterior ventral plate opisthosomal venter apodeme IIT apodeme IV apodeme V posterior median apodeme axillaries I axillaries IT internal and external presternals internal and external poststernals opisthosomal ventrals dorsals I laterals I dorsals IT dorsals III laterals IIT dorsals IV laterals [V internal caudals external caudals I external caudals II 5S THE UNIVERSITY SCIENCE BULLETIN coxa IV (or from near the lateral terminations of apodemes IV in some Pyemotinae). They may migrate anteriorly as far as the center of the third coxal foramen (some Microdispini). They are invariably present. The internal poststernals arise in line with or behind the second axillaries and close to the posterior median apodeme. They may be absent in some Pyemotinae and Pygme- phorinae. The external poststernals arise near or (usually ) posterior to a line drawn between the anteromesal margins of coxal foramina IV, and are distinctly lateral to the internal poststernals in all known species. They may be absent in some species. Internal morphology. The sclerotized portions of two systems, the digestive and reproductive, may be recognized in the hystero- somal region of cleared specimens of pyemotid mites. In unfed mites the former may be seen only rarely posterior to the dilated esophageal pumps. In such cases, it appears as a straight tube, similar to the esophagus in diameter, which seemingly ends blindly in the region of the third tergum. Several authors have thought the anus to be present but it has been established, at least in Pyemotes (vide Brucker, 1901, and Pitau, 1936) and Siteroptes (vide Reuter, 1909a), that the gut ends blindly. Reuter states that the anus is lost completely in Siteroptes except for a transitory “Exkretionsporus’ found in the early embryo. The structure of the reproductive system is much more complex. Brucker (1901) and Reuter (1909a) have investigated this system in Pyemotes and Siteroptes respectively and report similar findings which may be summarized as follows: The female genital organs are simple and consist of the vulva, which serves both as the true vulva and the ovipore or birth opening. This opens into the tubular vagina (oviduct of Reuter), from which a pouchlike dilation (the receptaculum seminis ) may arise. The vagina opens anteriorly into the single bulbose ovary. In both of the above species the female becomes greatly swollen when gravid. Information presented below indicates that, at least in the more specialized pyemotids, the above anatomical details are probably oversimplified and that a separate oviporus is present in at least some forms. Certain sclerites or genital thickenings found in cleared specimens have been largely ignored by earlier authors. The functions of these structures are not presently known but their presence and shape are useful in classification. In the Pyemotinae (except for some Siteroptoides ) and Acarophenacinae, such sclerites are usually indistinct, except that rodlike, globular, or omega-shaped structures THe Famity PYEMOTIDAE 59 are often found in Pyemotes, Trochometridium, and Siteroptes. These latter structures have been thought to be spermatophores (Pitau, 1936) or spermathecae. Species of Siteroptes and Trocho- metridium possess, in addition, a subapical, circumvaginal sclerite which may be subglobose to elliptical in shape. Members of the Pygmephorini possess a variety of distinct scle- rites in addition to this vaginal sclerite. In the Microdispini and Neopygmephorini, and in some Pygmephorini, one or more of three genital sclerites may be present; these are probably homologous be- tween species. For the sake of convenience, they are designated simply as the anterior, median, and posterior sclerites. These sclerites vary somewhat in shape and distinctness between species and even between specimens of the same species (due to mounting ) yet are usually recognizable and useful in classification. The anterior genital sclerite is usually long-triangulate, some- times appearing to be folded flaplike, its apex directed anteriorly. It arises near a line drawn between the posteromesal margins of coxal foramina IV except in Pygmephorus and some Pediculaster in which it arises well behind this line. The posterior genital sclerite is situated well behind the anterior genital sclerite and anterior to any apical vaginal sclerotization or structure. It is usually well defined and elongate straplike, or triangular. If triangular, it is also folded, flaplike, and seemingly arises from the anteroventral margin of the cuplike fifth segment. In the more primitive Pygmephorini this sclerite is usually not separable from several accompanying ones and may be absent en- tirely. The median genital sclerite (fig. 91) is a round, triangular, or irregularly shaped structure situated between the two preceding sclerites. It is present only in Acinogaster and in some Parapygme- phorus. Structural modifications associated with physogastry. Physo- gastry, the monstrous enlargement of the body of gravid female mites, has been previously thought to occur throughout the family Pyemotidae to about the degree well known in the genus Pyemotes. Although much work remains to be done on this problem, seem- ingly the extent of such physogastry varies greatly among the genera of the family. Moreover, the areas of the body entering into physo- gastry differ quite markedly in different genera. Genera known to swell to several hundred times their original volume include all of the genera of the Pyemotinae and of the 60 THE UNIVERSITY SCIENCE BULLETIN Acarophenacinae except Caraboacarus and Adactylidium (about which information is not available), and at least some species of Pediculaster and Perperipes. The latter genus is of particular con- cern here and will be mentioned later. Females of Pygmephorus are known to swell considerably but seemingly to a lesser extent than the genera mentioned above. The same is probably true for Microdispodides and some Parapygme- phorus, although specific information concerning them is not avail- able. As regards the Microdispini, all gravid specimens of Micro- dispus and Peponocara studied by this author contained only one or two large eggs and showed only a slight increase in body size. This situation is nearly identical to that occurring in the genus Imparipes of the family Scutacaridae. On the other hand, Per- peripes, which is considered to be one of the most specialized forms within the Microdispini, becomes enormously physogastric. How- ever, its physogastry is believed to differ considerably from that found in the Pyemotinae (see below). In the groups exhibiting pronounced physogastry, three main types of this condition, based upon the location and extent of the areas involved, may be recognized. (1) Opisthosomal physogastry. In this type the metapodosoma enters only slightly into the swelling and pleated conjunctival bands between the propodosoma and hysterosoma are absent or weakly developed. Swelling therefore occurs only in the opisthosoma, this region becoming enormously distended and spherical. As far as known, physogastry of this kind occurs only in the genus Pyemotes. (2) Hysterosomal physogastry. This is the most common type found within the family, the entire hysterosoma becoming swollen and spherical. A distinct band of pleated conjunctiva is usually present between the propodosoma and hysterosoma. This type of physogastry is known to occur in Dolichocybe, Siteroptes, Trocho- metridium, Resinacarus, Pediculaster, and, to a lesser extent, in Pygmephorus. (3) Idiosomal physogastry. This type is restricted to the more specialized members of the Acarophenacinae (Acarophenax, Para- carophenax, and probably Adactylidium ) in which the propodosoma is bipartite in the manner described previously (p. 49) and there- fore enters into the swelling. The entire body, except for the greatly reduced gnathosomal portion (including the arcula of the propodosoma) is therefore affected. THE FAMILY PYEMOTIDAE 61 In the physogastric females thus far examined, it is obvious that most of the gross enlargement of the body which accompanies the development of the ova takes place in the intersegmental conjunc- tiva, since measurements of the sclerites indicates that these do not change appreciably in size. What proportion of this conjunctival enlargement is due, however, to the unfolding of the conjunctival pleats and what proportion to conjunctival elasticity is not known. Richards (1951) pointed out that cuticle varies considerably in this regard, elasticity being correlated with the degree of orientation of the micelles, and it is possible (as in Ixodides) that considerably more elasticity occurs in the cuticle than has been heretofore sup- posed. Physogastry in Perperipes. The physogastric females (figs. 8, 9) of this genus differ greatly from those of all other pyemotids, and apparently are mimics of the doryline ant larvae among which they live. In these females the pleats of the intersegmental membrane vary in diameter in such a manner that, posterior to the second tergum, they form a regular series of 7 to 9 annuli when extended, these superficially resembling the segments of a larval ant. The fact that these “pseudosegments” may possess spicules like those of maggots leads one to suspect that this mimicry may be further extended to include locomotory or other behavioral characteristics. During physogastry the female becomes increasingly elongate, although body width remains approximately the same as in the non-gravid form. This elongation takes place only in the following regions: between dorsal segments II and III, between III and IV, and between IV and V. Only a slight separation occurs dorsally between the propodosoma and hysterosoma. Dorsal segments I and II bear the same relationship to each other as in the non-gravid female. Ventrally, the female is unmodified as far posteriorly as the hysterosomal venter, the latter remaining in its original position at all times. From beneath this sclerite, however, and apparently attached to the hind margin of the posterior ventral plate, there arises a greatly elongate conjunctival area which extends to the posterior end of the venter, where it joins the fifth segment. This area is easily visible in cleared specimens of non-gravid females as a thick band of conjunctival pleats. In physogastric specimens, its anterior portion, located directly behind legs IV and beneath the bysterosomal sternum, is distinctly convex and has a longitudinal 4— 1367 62 THE UNIVERSITY SCIENCE BULLETIN slit, the oviporus, along nearly its entire length. This convex por- tion, herein designated as the oviporal plate (fig. 8), is darker and more brittle than the more posterior portion of the ventral con- junctiva and is separated from it by an apodemelike ring which is not distinct in all specimens. This ring occurs ventrally in the region of the posterior genital sclerite and joins the posterior margin of the second hysterosomal tergum dorsally, thus, in effect, dividing the entire mite into two areas. The anterior of these areas is bulbose and slightly larger in girth than the posterior and, because it is composed nearly entirely of sclerotized plates, appears somewhat darker than the latter. It superficially resembles the head of a larval coleopteran or nema- tocerous fly. The posterior portion is composed almost entirely of conjunctiva except for the widely separated terga III to V. As mentioned above, this conjunctiva is constricted at regular intervals at and between the true segments, giving the female a segmented appearance. Numbers of these “pseudosegments” appearing be- tween the true segments are as follows: two or three between segments II and III; three between III and IV; one or two between IV and V. Segment V is always situated apicoventrally or ventrally. The origin of the oviporus is unclear. Its anterior termination is seemingly marked by the anterior genital sclerite, its posterior termination by the posterior genital sclerite. It is not visible in non-gravid females. It is probably either a modification brought about by and unique to the gravid state or, as seems more likely, is always present in the pleated conjunctiva of non-gravid forms, becoming definitive during the unfolding of the latter. Its length is such that it would extend beyond the hind margin of the body in non-gravid forms. Its elongation is brought about by the posterior migration of the posterior genital sclerite, the anterior sclerite re- maining in the same position as in the non-gravid form. Musculature in Perperipes. Two sets of longitudinal muscles are seen to pass through the first tergum to connect the posterior dorsum of the propodosoma with tergum II. Terga II, III, and IV are connected to each other by two similar sets of muscles, these capable of elongation to permit the expansion of the body in these areas. A single pair of muscles connects the fourth tergum with segment V. All longitudinal muscles arise anterior to the setal bases. In addition to the longitudinals, dorsoventral muscles are present as follows: (1) three pairs between the first tergum and the venter, the anterior pair originating upon the posteroventral margin of the THe Famity PYEMOTIDAE 63 anterior ventral plate, the mesal pair upon the lateral prolongation of apodemes III, and the posterior pair upon the posterior ventral plate anterior to coxae III; (2) one pair between the second tergum and coxae III; and (3) one elongate pair between the third tergum and the oviporal plate, immediately behind the posterior genital sclerite. No dorsoventral muscles were seen between tergum IV and the opisthosomal venter or segment V. If these muscles are indicative of the primitive relationships between the terga and sterna of true segments, it is clear that at least the ventral division between propodosoma and hysterosoma is of a secondary nature. As regards the relationship between the third tergum and the oviporal plate, it is possible that this plate represents the invaginated third sternum, the opisthosomal sternum therefore representing the sternum of the fourth segment. In view of Mitchell’s (1955) findings in the Unionicolidae as well as the observations mentioned earlier (p. 47) concerning the stability of muscle attachments, any speculation based largely or entirely upon these would seem questionable. Reproductive organs of Perperipes. The most striking feature of this system and one not visible in non-gravid females is the large, muscular oviduct, whose external opening is the oviporus. Within this duct fully formed larvae may be seen, these still within the chorion. More or less within the center of the greatly enlarged uterus and in direct access to the oviduct are the most mature em- bryos, these decreasing in maturity as one proceeds posteriorly. On each side of the uterus and lateral to the oviduct a germarium is present, the embryos showing a progressive increase in maturity as they move from this region posteriorly to the tip of the hystero- soma. In the most posterior portion of the uterus the embryos from each side merge to form a single stream moving anteriorly into the oviduct. The true vulva, still visible in segment V, appears to be dis- sociated from the uterus. It clearly does not function in oviposition. Legs. The legs of pyemotid mites may be composed of either four or five segments. As explained earlier, this author has chosen to call the most basal free segment the coxa, this segment therefore corresponding to that designated as trochanter or trochantofemur by most European workers. The next proximal segment is weakly to strongly subdivided, this being most noticeable on legs III and IV. This segment is called the trochanter herein. The remaining segments, femurogenu, tibia, and tarsus, are named in order begin- 64 THE UNIVERSITY SCIENCE BULLETIN ning with the most basal, their nomenclature corresponding with that in use by most European acarologists. The pretarsus is not considered herein to be a segment. In cases where one segment is lost, the loss is due without exception to the fusion of tibia and tarsus. This fusion occurs in leg I of all Acarophenacinae but Cara- boacarus, in the genus Resinacarus, and in all Pygmephorinae. Caraboacarus is unique in having tibia and tarsus IV fused, the line of fusion still visible. The articulation of the legs (notably that of I and IV) to the body undergoes an interesting evolution within the different genera. In the Pyemotinae and in the less specialized Pygmephorinae (some Pygmephorus, some Pediculaster ) legs I are carried straight forward and their articulations (as seen in ventral aspect) are at approxi- mately a right angle to the longitudinal axis of the body. Within the Acarophenacinae and Pygmephorinae, a parallel transition to an oblique type of articulation occurs, this transition seemingly quite abrupt in the former group. Whether the change in the carriage of this pair of legs is correlated with a corresponding change in their function is not known. In the more primitive Pyemotinae, e. g., Dolichocybe, Pavania, Pyemotes, Metasiteroptes, as well as in most Acarophenacinae, coxae IV are triangular and similar to coxae III in shape and points of articulation, and therefore in movement. In these groups, two ventral condyles, one at the anterior, one at the posterior extremity of the coxa permit the entire leg to be abducted and adducted at approximately a right angle to the longitudinal axis of the body. Flexion-extension is accomplished at the coxo-trochanteral joint. In the higher Pyemotinae and in the entire Pygmephorinae, coxa IV differs from coxa III greatly in shape, number and position of con- dyles, and in function. In these mites the coxa becomes quadrate (Siteroptes, Trochometridium), then rectangulate and greatly flat- tened (most Pygmephorinae). In the latter, the coxal foramen is nearly slitlike and the base of the coxa becomes increasingly cov- ered by the posterior ventral plate. Concurrently, there is a gradual posterior migration and accompanying enlargement of the anterior sternocoxal condyle until, in the Pygmephorinae, this condyle is found on the inner surface of the posterior ventral plate near the center of the coxal base, where it projects dorsally into a socket on the ventral surface of the latter. In some but not all Pygme- phorini the posterior condyle seemingly is vestigial or lost. It ap- pears that in these forms the capability for abduction-adduction THe Famity PYEMOTIDAE 65 of legs IV has become restricted and may be, in some cases, par- tially replaced by one for flexion-extension. In this respect, these mites show a close affinity to those of the family Scutacaridae. Tarsi I to IV vary in the number and development of the claws, these oftentimes being of diagnostic value. Tarsus I bears zero to two, the latter condition being present only in the genera Pavania and Dolichocybe, the former having arisen independently in a number of genera. The usual number of claws on this leg is one and the pulvillus is lacking except in the form designated herein as Dolichocybe fusiformis (Krezal, 1959), and in Pavania. Claw I may be sessile or pedicellate in varying degrees. It is usually en- closed basally within a membranous or sclerotized cuplike pretarsus which may be elongate along its inner margin to form an opposable “thumb.” The latter may be absent (Siteroptes, most Pseudo- pygmephorus) to well developed (most Pygmephorus ) and may be entire, dentate, or cleft for the reception of the claw. In Trocho- metridium, two thumbs are present, one of which is setalike and perhaps is not of pretarsal origin. This thumb may be capable of movement. In many forms in which claw I is sessile, its position is twisted or otherwise distorted. The pedicel is always present and usually long and slender in those forms in which the thumb is absent. Claw I may be tiny and subsetiform (some Myrmecodispus ) to very large (Pygmephorus) and it may be simple and _hooklike (Siteroptes, some Pseudopygmephorus) or bulbose basally (many genera). A general correlation exists between the size and shape of claws I and the occurrence of phoresy. Notable exceptions occur in Caraboacarus, Peponocara, and Glyphidomastax, in which the claws are lacking. Phoresy is only rarely known in species in which claws I are long-pedicellate or simple and hooklike. The ambulacra of legs II and III are invariably identical in shape, both claws and pulvilli being present except in Acarophenax and Adactylidum, in which the claws are lacking. If present, the claws of these legs are always paired and may be simple, cleft, or toothed. The pulvillus is disclike or flaplike, usually longer than the claws. Tarsus IV resembles tarsus I in having lost the claws independently in several genera (Caraboacarus, Acarophenax, Adactylidium, Mi- crodispodides, Glyphidomastax, Perperipes) but differs in retaining the pulvillus after such a loss. Claws IV, if present, are often re- duced in size and are only rarely cleft or dentate. 66 THe UNIVERSITY SCIENCE BULLETIN In several genera, a round or irregularly-shaped thickened area may be present on tibiotarsus I. This area is mediolateral in loca- tion, usually arising between solenidia 2 and 3, and perhaps is as- sociated with the fusion of the tibia and tarsus. It may appear dorsal, circular, and well separated from the lateral margin of the segment, or marginal and irregular. Setation. The number and placement of leg setae are useful in classification at both the specific and higher category levels. Krezal (1959b) has mapped and attempted to homologize these setae. The practical benefit derived therefrom is great and the setal nomen- clature proposed by Krezal is used in this paper (figs. 10-12). It should be pointed out, however, that considerable shifting of these setae may occur, and in addition certain tactiles may be replaced by sensory setae and vice versa. The identification of homologues must therefore be viewed with reservations. Coxae I to IV always bear one seta except in Dolichocybe, Pavania, and in some Adactylidium, in which coxal setae are lack- ing. The setation of the remaining segments differs considerably between legs and is best discussed by reviewing the latter in sequence. Trochanter I bears four setae in the Pyemotinae except for Trochometridium (which has five), in the higher Acarophenacinae, in Pediculaster, and in some Pygmephorus, all other forms bearing three. The structure of seta c (the apicolateral dorsal) of this seg- ment is variable, being setiform in the Pyemotinae, Acarophenaci- nae, and Microdispini, but stout, elongate, and obtuse or spatulate apically in Pediculaster, and short, bladelike, and usually proflexed apically in the remaining genera. This same seta undergoes further modification in the family Scutacaridae (Beer and Cross, 1960). Seta d (the apicomesal dorsal) of this segment may be setiform or modified and undulate. Trochanter II usually bears three setae (two in Dolichocybe), trochanter III] from one to three, and tro- chanter IV two or, rarely, one. Femurogenu I bears three or four tactiles except in Dolichocybe and Pavania, which have only two. Femurogenu II possesses from one to three, femurogenu III from one to three but usually one less than the number found on femurogenu I, and femurogenu IV one or two. Tibia I, if separate, bears five or six tactiles and two solenidia (nos. 1 and 2 of Krezal, 1958). Tibiae II to IV invariably bear four THE FAMILY PYEMOTIDAE 67 tactiles and zero, one, or two solenidia in addition. The usual num- ber of such solenidia is one (Solenidium 1 of the leg concerned ) and if two are present (only on tibia II of some Parapygmephorus ) then one arises from the apical margin of the segment and is un- doubtedly the basal tarsal solenidium (Solenidium 2). The single true solenidium (1) of the tibia becomes lost more often from leg III than from leg II and from leg IV more often than from leg III. Genera lacking this solenidium on leg III are Paracarophenax and some Pyemotes, while the same solenidium is absent from both legs II and III of the genera Dolichocybe, Acarophenax, and Adactyli- dium. Tarsus I, if distinct, bears 13 to 15 setae, six to nine of which are tactile. Tarsus II bears from five to eight tactiles and a single solenidium which is invariably present except in some Parapygme- phorus where it may arise from the tibia (above). Tarsus III bears from four to eight tactiles and always lacks a solenidium. Tarsus IV bears from three (Adactylidium) to seven (Trochometridium) tactiles, the usual number being six. Solenidia are always absent on tarsus IV. In the subfamilies Pygmephorinae and Acarophenacinae except for Caraboacarus, and in the pyemotine genus Resinacarus, tibia and tarsus I are fused. In these genera the total number of setae varies between 17 and 21. The two short, lateral, usually clavate solenidia found on each of the two original segments (solenidia 1 to 4) are easily recognizable and provide landmarks for further setal recognition. In the Pygmephorinae and in the family Scuta- caridae, certain sensory setae may be borne upon distinctive cylin- drical or irregularly-shaped pedicels termed pinnacula herein. The number, shape, and setal complement of these structures may be useful in classification at specific or generic levels. The setal designations of Krezal probably oversimplify the true setal picture of this compound segment since, in reality, about half the setae, particularly those of the lateral surface, are rodlike sensory setae, many so closely resembling tactiles as to be nearly indis- tinguishable from them. It is obvious that use of such terms as “tactiles” is quite arbitrary and merely follows custom; nothing is known of the functions of the various sensilla. 6S THe UNIVERSITY SCIENCE BULLETIN MORPHOLOGY OF THE MALE General As in the remaining Tarsonemoidea, a great deal of dimorphism occurs between the sexes. Unlike the remaining families, however, in which this dimorphism is usually uniformly established to a high degree, males of the Pyemotidae show a sequence of modification from which one may speculate upon probable relationships occur- ring within the superfamily. This sequence begins with Dolichocybe and terminates along one phyletic line with the Neopygmephorini and the Scutacaridae, males of these two being indistinguishable from each other at the present time. A second line of descent, less clear than the preced- ing, seemingly passes from a Dolichocybe-like form to the Acaro- phenacinae. Mites of this subfamily show many affinities to the Tarsonemidae and the Podapolipodidae but their relationship to these two families is not clear. Although Dolichocybe is considered here to be primitive, even in this form a great deal of morphological disparity occurs between the sexes. In addition to primary sex modifications, this dimorphism involves body shape (the female being elongate and cylindrical, the male much shorter and stouter), gnathosomal shape (more elongate in females), the length of the legs (comparatively longer in the female), the loss of the pseudostigmata in the male (a pair of setae is present which are probably their homolgues, however ), and very minor setal differences on the legs. When compared with the extreme modifications of males in other genera, however, the males of this genus are relatively unspecialized. Because of the lack of material, males being rare in many species, a more detailed study of the comparative morphology of the males of the family is unfortunately not possible at this time. The fol- lowing is based on information obtained from the study of the males of the following: Dolichocybe keiferi; Pyemotes (one species); Siteroptes, s. str. (four species); Trochometridium tribulatum; Neo- pygmephorus, s. str. (one species). In addition, males of several unidentified species of Pygmephorinae as well as of the families Scutacaridae and Tarsonemidae were also utilized for comparative studies. The drawings and descriptions of the males of Acarophenax, Paracarophenax, and Resinacarus (found respectively in Newstead & Duvall 1918, Rack 1959, and Vitzthum 1927) have provided addi- tional information. THe Famity PYEMOTIDAE 69 As in the females, taxonomic separation of the males of the more primitive genera is seemingly quite simple, that of higher genera less so. Perhaps, in the case of the males, this is illusory because of the few species available for comparison. Gnathosoma This region may be functional and developed to varying degrees (Dolichocybe, Pyemotes, Resinacarus, Acarophenax, Paracarophe- nax, all Tarsonemidae and Podapolipodidae) or completely reduced and nonfunctional (Trochometridium, Siteroptes, Neopygmephorus, two other unidentified Pygmephorinae, and all known Scutacaridae ). In the latter case, this region consists only of one or two lobes bear- ing several pairs of setae. The oral opening, mouthparts, and palps are completely lacking when the gnathosoma is reduced. In forms in which the gnathosoma is well developed (Pyemotes, Dolichocybe), its setation is similar to that of the female. In the latter genus the gnathosoma is shorter and broader than that of the female and differs otherwise in having the palps fused to the gnatho- soma. The pharyngeal pump and esophagus is distinct in males of this genus despite the fact that the chelicerae are not visible. Propodosoma This region is always wider in the male than in the female. The dorsum of males differs from all females except those of Dolichocybe and Pavania in lacking peritremes, and indeed, all traces of stigmata. Males further differ from all females, except those of the specialized Acarophenacinae, in lacking pseudostigmata. In all genera ob- served except Trochometridium, however, an extra (fourth) pair of propodosomal dorsals occur which may well be the homologues of these organs. The propodosomal venter resembles that of the female in setal number except in the Pygmephorinae. Generally speaking, the apodemes of the anterior ventral plate resemble those of the female. Males of Neopygmephorus differ from their females in that their first coxae articulate in the primitive manner, i. e., at a right angle to the longitudinal axis of the body. In all other genera studied the sexes agree in this respect. The posterior margin of the male propodosoma usually meets, or is separated by a wide zone of conjunctiva (Trochometridium ) from, the hysterosoma. It neither covers nor is covered by the an- terior margin of the first hysterosomal tergum. 70 THe UNIVERSITY SCIENCE BULLETIN Hysterosoma Like that of the female, the male hysterosoma is capable of undergoing great modification in many, if not most, areas and organs. These modifications generally seem to be adaptively cor- related with the use and disuse or modified use of the legs and genitalia. Only in Dolichocybe does the male retain a close resemblance to the female in numbers of segments and setae. In this genus four bandlike dorsal segments are visible, the fourth apparently covering a reduced fifth which may be fused with the opisthosomal venter. The number of dorsal setae resembles that of the female. In all remaining genera a fusion of terga I and II has occurred, the maxi- mum number of dorsal plates therefore being three. In Paracaro- phenax and Resinacarus tergum III appears to be terminal, but in the remaining genera this sclerite is subterminal and platelike. It bears one (rarely) or the usual two pairs of setae. The males of Acarophenax, Paracarophenax, Dolichocybe, Trochometridium, and Resinacarus are broadly rounded or broadly truncate posteriorly. In these forms the opisthosoma is not greatly constricted or otherwise modified into a globular or conical copulatory structure. In the first two genera above, the aedeagus is displaced ventrally and emerges from a genital plate which may consist of portions of tergum IV and segment V. In Dolichocybe and in Trochometridium the genital aperture is ventral and terminal but the latter genus differs from the former in that a terminal concavity exists between segment IV and segment V, the latter bearing the aedeagus. In Pyemotes, Siteroptes, Neopygmephorus, and in at least some Scutacaridae, tergum IV, segment V, the opisthosomal venter, and perhaps some phallic structures associate to form a conelike or sub- globular genital capsule. Since the gnathosoma in these forms is vestigial and because the aedeagus is prominent and often bipar- tite, this capsule easily may be mistaken for the gnathosoma. The genital capsule may be relatively small and subtruncate or concave apically (Pyemotes) to very large and acuminate (Neopygme- phorus). Tergum IV forms the dorsal plate of this capsule. This tergum is seemingly modified for the reception of the female opis- thosoma during copulation and may be roughened, punctate, con- cave, or provided with membranous “wings.” It usually bears two pairs of basal setae, of which at least one pair is sensory. Apically, tergum IV encloses the fused fifth segment and opisthosomal venter, THE Famity PYEMOTIDAE gat these last containing the male genital orifice. At least in Neo- pygmephorus, these last two sclerites may be tripartite, the two lateral lobes formed by such a division being flaplike. It is possible that these flaps may further assist in positioning the female prior to copulation. The basal portion of the genital capsule is enclosed ventrally by the greatly constricted hind portion of the posterior ventral plate except in Pyemotes, in which the ventral portion of the capsule may be free. Broadly speaking, the arrangement and development of apodemes III and IV found in Dolichocybe and Acarophenax resemble that of their respective females. In Paracarophenax these apodemes are distinctly stronger than in the female. In the remaining genera they are strong, usually complete, and often differ markedly in location and shape from those of their respective females. These changes are probably associated with the modification of legs IV as clasping organs and may reflect corresponding changes in the copulatory position. In Dolichocybe, Pyemotes, Acarophenax, Paracarophenax, and Trochometridium, the penis is a short, rigid, blunt or hooked organ which is usually divided longitudinally into halves. In Siteroptes it is longer, tubular, and acuminate apically but still rigid. The males of Neopygmephorus and the scutacarid genus Imparipes possess a structure similar to that of Siteroptes (fig. 100) but very long and flexible and having taenidia-like thickenings. The author has studied the internal reproductive structures of Siteroptes (Siteroptes) reniformis Krantz and of an undescribed Neopygmephorus. In these species a duct, the vas deferens, con- nects the single testis to a globular chamber which probably func- tions as a combined seminal vesicle and sperm pump. This chamber is characterized by lateral winglike structures which are probably dilator muscles. The tubular ejaculatory duct connects the pump with the intromittent organ. In S. reniformis, while this duct ap- pears superficially to open ventrally into the penis about halfway along the latter’s length, closer inspection reveals that it doubles back to join the penis at its base, therefore allowing protrusion of that organ without imposing strain on the duct. In Neopygme- phorus and Imparipes this reflexed portion is lacking and the ejacu- latory duct is seen to enter the penial base. This arrangement is in keeping with the nature of the intromittent organ, which in these genera is long, flexible, and looped within the genital capsule. ey THE UNIVERSITY SCIENCE BULLETIN Legs The legs of male pyemotids may be quite similar to those of their respective females or greatly modified. Generally speaking, the similarities are most pronounced in Dolichocybe, least in Neopygme- phorus. As in the females, and with but few exceptions, legs II and III are the principal locomotory appendages and are therefore thought to retain the least specialized form and setation. Leg I invariably possess five segments, the setation of the tibia and tarsus closely following that believed by Krezal (1959b) to be primitive (fig. 12). In Dolichocybe tarsus I bears two claws, in Trochometridium it bears only a membranous flap, and in all re- maining known forms, a single claw is present. Trochanter I re- sembles that of the female as regards setal number but setae c and d are always setiform. The angle of articulation of legs I to the anterior ventral plate is the same in both sexes except in Neopygmephorus and the scuta- carid genus Imparipes. In these genera coxa I of the male articu- lates at a right angle to the longitudinal plane of the body, while in the female it is distinctly oblique. The relative length of the legs may differ greatly between sexes. In the male of Neopygmephorus, for example, leg I is much longer than leg II, while in the female the reverse is true. Leg II of Trochometridium is distinctly larger than leg HI. The setation, although similar to that of the female in number, differs distinctly in structure. These changes are probably correlated with the mating behavior of the male (see p. 126). In Resinacarus, leg HI is noticeably longer than II and has an elongate, spurlike seta lacking on the former. According to the original description (Vitzthum, 1927), only this leg differs greatly from its female counterpart. Leg IV closely resembles that of the female only in Resinacarus and in Dolichocybe. It is distinctly stouter and shorter than its female homologue in Acarophenax, Paracarophenax, and Trocho- metridium and usually has at least one modified claw in addition. In the male of Paracarophenax dermestidarum (Rack) legs IV are distinctly larger than legs III, but the reverse is true in Acarophenax tribolii. Legs IV are modified forcepslike in all remaining genera studied and are normally carried horizontally and directed toward the rear. Concomitant modifications may occur in the shape and size of the coxa or trochanter IV or both, a shortening of the tarsus, and the THe Famity PYEMOTIDAE 73 simultaneous enlargement of one claw and dimunition or loss of the other. In addition, an accompanying increase in the strength of the posterior ventral apodemes usually takes place. Setation in these forms is similar to that of the female except that one or two tibial setae may be spinelike and that the tarsus may be lacking a seta. IMMATURE STAGES Insofar as known, members of the genera Dolichocybe, Acaro- phenax, Paracarophenax, Pyemotes, and Trochometridium possess no active immature forms, the adult mite being the only definitive postoval stage. Two free-living immature stages, the hexapod larva and one octapod nymph, have been reported in Siteroptes graminum (vide Reuter 1900, Cooper 1937, Alfaro 1946), but the appearance of the latter should probably be investigated more thoroughly since it seems likely that the “nymphs” are in reality young females. Fully developed larvae still within the chorion have been observed by the present author in mounted specimens of Perperipes and Neopygmephorus. In at least one species of the latter genus, the larval stage is not free and the nymphal stage is lacking, since the adult mite may be seen within the skin of the molting larva. This developmental sequence of egg — larva > adult also occurs in at least two species of the scutacarid genus Imparipes (G. Bohart and Cross unpublished), although in the latter genus the larval stage is free. The larvae of two species of Siteroptes and a single species each of Perperipes and Neopygmephorus agree more closely with their males than with their females as concerns the configuration of the anterior ventral plate and the structure of the legs. In Siteroptes, however, larvae possess only four pairs of anterior ventral setae while both males and females have six pairs. Larvae of the above three genera further resemble males in lacking tracheae, peritremes, the pseudostigmata, and in possessing an extra pair of propodosomal dorsals which are probably homologues of the latter. Three hysterosomal terga are present and of normal size but the fourth and fifth are reduced and modified to form a pronounced terminal lobe. Setation of the first four segments resembles that of the female. The fifth segment bears at least two pairs of setae. Although the anterior ventral plate closely resembles that found in the male, the setation and apodemes of the posterior ventral plate differ considerably from those of both adults as might be expected, due to the loss of a pair of legs from this plate. 74 THe UNIVERSITY SCIENCE BULLETIN Leg I is always five-segmented and is articulated to the anterior ventral plate at a right angle to the longitudinal axis of the body. As in the male, setae c and d of trochanter I are always setiform. Tibia and tarsus I resemble these segments of the male except that the latter always has two claws. Legs II and III likewise resemble those of the male. PHYLOGENY In view of the fragmentary knowledge of the Pyemotidae and related groups, this portion of the present work is preliminary. It is included as a beginning from which further speculation may de- velop. The construction of a scheme of phylogeny properly supposes a knowledge of ancestral forms, but this is not possible in the Tarso- nemoidea owing to the complete lack of a fossil record. Specula- tions concerning the relationships of the groups were therefore made by less satisfactory means. The genera were first arranged into higher categories based upon the largest number of characters possible. These characters may be found in Tables III and IV, in the keys, and in the descrip- tions. Estimates of higher categories which were most closely re- lated were then made by noting the intermediate genera, i. e., genera in one tribe or subfamily sharing the largest number of characters with another tribe or subfamily. Such estimates are probably con- sciously or subconsciously a part of the procedure followed by most systematists and are usually used to deduce the supposed phylogenetic relationships. Static or phenetic relationships may be much more precisely defined in terms of correlation coefficients among species as done by Michener and Sokal (1957). Characters were considered to be of two kinds with respect to their usefulness in determining relationship: (1) Those for which no assumptions of primitiveness or specialization were made, and (2) those for which inferences of primitiveness or specialization were made. The former are invariably characters arising within the group or characters whose homologues outside the Tarsonemoidea are uncertain. For example, a single pair of opisthosomal ventral setae is present in a few Pyemotinae, in all Acarophenacinae and Tarsonemidae, but is absent in the remaining Pyemotinae and in all Pygmephorinae and Scutacaridae. The homologues of these setae may or may not be any of a number of similarly located setae found in the groups considered to be most closely related to the THE FamMIty PYEMOTIDAE 75 Tarsonemoidea, hence, by comparison cannot be considered to be either primitive or derived. Such characters are nevertheless valid in evaluating phenetic relationships. In a few cases (Table III, males only, and Table IV), particularly when they formed sequences of modification, they were used in association with other characters in the phylogenetic speculation. Inferences as to the degree of primitiveness or specialization of characters thought to be useful in making phylogenetic assumptions were reached as indicated below: (1) By comparison of the various tarsonemoid taxa with non- tarsonemoids considered to be closely related. For reasons to be mentioned later, the Anystoidea of Cunliffe (1955) are probably more closely related to tarsonemoids than are any other mites and are ancestral to them. The pyemotids of the Tarsonemoidea are most similar to the pseudocheylids of the Anystoidea. Comparisons of characters between members of these two superfamilies, and particularly between the two families mentioned, were considered to be more meaningful than similar comparisons made with groups thought to be more distantly related. Characters found in the anystoids but only some tarsonemoids are considered to be more primitive in the Tarsonemoidea than the alternative characters found only in the Tarsonemoidea. These comparisons present, however, several important difficulties, as follows: (a) Uncertainty does still exist as to the closest relatives of the Tarsonemoidea. The gap between tarsonemoids and anystoids is of sufficient size that explicit lines of descent cannot be recognized. {b) Any repeti- tive character (i.e., one lost or modified, then regained) or any character that arose independently in different groups may produce similarities resulting from convergence or parallelism rather than phylogenetic relationship. Despite these difficulties, this method of recognizing primitive and derived characters is probably the most reliable and has been used herein wherever possible. Table III lists characters judged on this basis. (2) By extension of the premise that phenetic similarity implies relationship. In certain cases where well-defined sequences of modification in the states (any one of the two-several alternative conditions possible) of a character were present, phenetic char- acters were used in the determination of phylogeny. In such cases, the sequences were present in two morphologically different lines as determined by numerous associated characters, and the primitive 76 THE UNIVERSITY SCIENCE BULLETIN end of the sequence was considered to be the end common to both. For example, a progressive change in the number of ventral propo- dosomal setae has occurred within the Pyemotidae. In most of the Pyemotinae six pairs occur but in the Pygmephorinae five and four pairs are the rule, and in the Acarophenacinae four, three, and two pairs are present. The latter number is characteristic of most Tarsonemidae (except for Tarsonemella). In this instance, the author has considered the primitive number to be six pairs. Char- acters whose primitiveness or specialization was determined in this fashion are listed separately in Table IV. (3) By a combination of methods 1 and 2 above. The primitive and derived states of the characters listed for males in Table III were arrived at by combining the two methods above. In this case, anystoid males were not available for study but infer- ences drawn from comparisons of tarsonemoids with prostigmate forms considered to be more distantly related than anystoids agree with those drawn from observations of sequential modifications, present only in males, in which one extreme contained forms closely resembling the females but the other extreme consisted of forms strikingly different from them. There is, for example, a progressive reduction of the mouthparts followed by the reduction, then loss of the gnathosoma. A parallel sequence concerns certain great modifications of the opisthosoma and copulatory apparatus. In this case, it seems clear that the primitive characters are those most nearly alike in both sexes. Single series of sequential modifications, unless supported by evidence from other sources as in the case of the males above, were not considered in the speculations on the phylogeny. In the Pyemotidae, certain characters are of special interest in that they appear highly specialized, yet occur only in the genera considered to be primitive. Extreme physogastry, for example, is found throughout the Pyemotinae, Acarophenacinae, and in most Pygmephorini. It is not known to be present in the Microdispini and Neopygmephorini (except in the microdispine genus Per- peripes) and in the Scutacaridae. A closely related character, the number of active immature stages, follows a similar pattern. In certain pyemotines (Dolichocybe, Trochometridium, Pyemotes) as well as in the Acarophenacinae, all immature stages are supressed, the first active stage therefore being the adult. In Siteroptes and in all known Pygmephorinae as well as in the Scutacaridae, a definitive hexapod larval stage is interposed between egg and adult. THE Famity PYEMOTIDAE 77 The occurrence of these unusual characters in forms believed closely related to the Pyemotidae has not been studied but partial physogastry in conjunction with the simultaneous production of at least eight eggs has been observed by the author in one gravid female tenerifhid. Despite their seemingly obvious specialization, such characters must be considered primitive within the family, and the return to the less extreme condition as derived. The same treatment is accorded characters which are lost, then regained, both loss and reappearance being considered as specializations. Groups having a high percentage of the primitive characters listed in Tables III and IV are considered primitive, those having a high percentage of derived characters are considered to be spe- cialized. Further judgment concerning the positions of the various taxa was made after consideration of habits and internal structures. Table V lists for each subfamily and tribe, in the manner of Sinha (1958), the number of primitive, partially derived (derived in some, but not all species in a group), and derived characters from Tables III and IV. Males are listed separately because they are so poorly known. Table VI indicates the percentage of primi- tive characters listed in Table III possessed by representative species of each genus. Males are listed separately for the reason stated above. Relationships of the Tarsonemoidea to other groups. Although at least two ideas regarding the relationships of the group have been popular in the past, all contemporary authors agree that the Tarsonemoidea should be placed in the Suborder Trombidi- formes and that its members are most closely related to the Super- cohort Prostigmata (see History). The recent discoveries of the primitive genera Dolichocybe and Pavania have, however, empha- sized a possible relationship, hitherto unsuspected, between the Pyemotinae and the anystoid family Pseudocheylidae. In general, anystoids resemble prostigmates and differ from tarsonemoids in that the cheliceral bases are freely developed and moveable, in possessing well-developed, segmented palps which usually retain the thumb-claw complex, in the position and size of the genital opening, the latter often furnished with suckers in the female, in genital setation, in usually having six or more leg segments, in lacking definite segmentation (although segmentally arranged plates may be present), in usually possessing eye spots, and because (usually?) the males resemble the females. Alternatives of some of these characters peculiar to the Tarsone- 78 THE UNIVERSITY SCIENCE BULLETIN TABLE III.—List of Primitive and Derived Characters in the Pyemotidae Primitive Derived A. FEMALES 1. Gnathosoma free Gnathosoma enclosed in propodosoma 2. Palps free Palps fused to gnathosoma 3. Palps not elongate, not fitted for Palps elongate, fitted for piercing piercing 4. Chelicerae small, distinct (a). Chelicerae reduced, not or barely visible (b). Chelicerae enlarged, bladelike 5. Stigmata present Stigmata absent 6. Pseudostigmata present Pseudostigmata absent pe 3 ig (a). in gnathosoma 7. Without well defined esophageal Hsophageal pump (). im aronedosomnn pump 8. Gnathosoma directed forward Gnathosoma directed at least antero- ventrally 9. Propodosomal dorsum just meet- Propodosomal dorsum distinctly over- ing first hysterosomal margin lapping or being overlapped by first : hysterosomal tergum. 10. Propodosoma not divided Propodosoma divided into two parts 11. Hind margin of posterior ventral | Hind margin of posterior ventral plate plate entire, plate lacking median | tripartite, plate with median triangular triangular area area 12. Posterior ventral plate whole Posterior ventral plate divided longi- tudinally or transversely 13. Anterior genital sclerite indistinct | Anterior genital sclerite distinct 14. Posterior genital sclerites indis- | Posterior genital sclerite triangular tinct or variously shaped but not | and flaplike triangular and flaplike 15. All legs with five segments Legs I or IV with four segments 16. All coxae similarly shaped At least coxae IV differing from the rest in shape 17. All tarsi with two claws (a). Tarsus I with only one claw (b). Claws lacking on one or more pairs of legs 18. Seta c of trochanter I setiform | Seta c of trochanter I modified 19. Tarsus IV similar to tarsi II and Tarsus IV more or less elongate, dis- Ill tinctly longer than II or III 20. Claws II-IV alike, simple Claws II and III basally dilate, often toothed in addition THe Famity PYEMOTIDAE 79 TaBLeE III.—List of Primitive and Derived Characters in the Pyemotidae—Concluded Primitive Derived A. FEeMALES—Concluded Coxa IV not markedly bulbose basally Mites pale, soft bodied B. MALEs Gnathosoma present, well devel- oped Mouthparts functional Genital capsule undeveloped Posterior ventral plate resembling that of female Dorsal segments I-II distinct Legs I-III similar to those of fe- male Legs IV straight, not greatly dif- fering from II and III in length Coxa IV slightly to markedly bulbose basally Mites sclerotized, harder bodied Gnathosoma absent or at least greatly reduced Mouthparts non-functional Genital capsule developed Posterior ventral plate differing from that of female Dorsal segments I-II fused Legs I-III differing from those of fe- male Legs IV tonglike, often greatly shortened TaBLE I[V.—Supplementary List of Primitive and Derived Characters in the Pyemotidae Primitive Derived tw With six pairs of ventral propodo- somal setae With three pairs of propodosomal dorsals Segment V with more setae Posterior ventral plate with six pairs of setae Trochanter I with four setae With fewer than six pairs of ventral propodosomal setae With fewer than three pairs of pro- podosomal dorsals Segment V with fewer setae Posterior ventral plate with fewer than six pairs of setae Trochanter I with three or (rarely) five setae SO THE UNIVERSITY SCIENCE BULLETIN moidea and believed to be derived are as follows: cheliceral bases are fused with the palpal bases to form a gnathosomal capsule in which only the moveable chela seems to function as a protrusible piercing organ, the palps are reduced with segmentation absent or indistinct, the distinct thumb-claw complex is lacking but a more or less palmate apical claw is often present, the female genital opening is apical, the legs possess five or fewer segments, and sexual dimorphism is present, usually marked. The presence of well-defined body segmentation in the Tarsone- moidea has sometimes been regarded as primitive. Although certain primitive mites of the Anactinochitinosi (e. g., Opilioacarus) retain what may be primitive segmentation, there is no account of such segmentation in the Trombidiform-Sarcoptiform group, and in view of their many other seemingly specialized characters, such segmenta- tion may well be secondarily acquired in the tarsonemoids. Pyemotines resemble pseudocheylids in being elongate and weakly sclerotized, in the shape and position of the peritremes (lost, how- ever, in Pavania and Dolichocybe) and of the pseudostigmata, in the shape, number of segments, type of pulvillus, and to a surprising degree, the setation of the legs, in lacking genital suckers, and in the setation of the propodosomal dorsum. Relationships among the families within the Tarsonemoidea. Broad trends of the relationships within the Tarsonemoidea are seemingly plain and were outlined, sketchily, as early as 1901 by Brucker. It is generally agreed that the Pyemotidae are the most generalized family, being most similar to prostigmata and anystoids, and give rise, on the one hand, to the Tarsonemidae and Podapo- lipodidae, and on the other to the Scutacaridae. The origins of the first two are obscure but their closest living relatives are seem- ingly found in members of the genus Pyemotes and in the subfamily Acarophenacinae (see below). The Scutacaridae, on the other hand, are doubtless specialized forms arising from the Pygme- phorinae of the Pyemotidae and are so closely related to them that their present family status may reasonably be questioned. Since females may be satisfactorily separated, and because such a separa- tion simplifies the classification, it is retained herein. Relationships within the family Pyemotidae. The family is here divided into the three subfamilies, the Pyemotinae, Acarophenaci- nae, and Pygmephorinae. Of these, the Pyemotinae has the small- est number of derived characters (Tables V and VI). All its mem- THe Famity PYEMOTIDAE 81 TasBLE V.—Number of Primitive and Derived Alternatives of Characters Listed in Tables III and IV Primitive Partly derived Derived FEMALES Ryemotimaes: 46-02 an 17 10 0 NCarophenacinae 1 e.cne a: 11 9 6 Pygmephorinae Pygmephorini............ 11 5 11 Microdispini............. 9 a 11 Neopygmephorini........ 10 3 14 MAatLEs Pyemotinae................ 1 6 0 Acarophenacinae............ 3 0 4 Ry cme phonimacne 2. ee aaele 0 0 iG Tas_e VI.—Per cent Age of Primitive Characters (From Table III) Possessed by Various Pyemotid Species Percent primitive characters SPECIES Male Female PYEMOTINAE IPO MG FUISHOUMUS 6S Skid ce ROA GS co abo eau alow ou OCS one Gaee 82 Dolichocybe keifert... 0.0.0.0... 0.00.0.0004. 100 96 TU CTOLESISP Nk ay i 56 76 IS CLEROMECSHINGCEN Re ace Mainicrs aa teasnne) “pone, Real ited Ge as ase 91 Siteroptes prob. graminum................ 0 91 Siteroptes (Siteroptoides) sp............... 0 86 Trochometridium tribulatum............... 44 76 ACAROPHENACINAE COR GDOGCOMUSES eee oh ae Mares oy ee a te eae oe (al FACOTOPMCTOTUnCOOULUM Wels nee aieicds We eanae 56 48 Paracarophenax dybast................... 56 48 Adacthy idan: beent ape teeerererice rancli yi Marcu einai 48 PYGMEPHORINAE Pediculaster americanum.................)00.0 000.000. 53 ZYONEPRLOTUS. (SPLNOSUS: CLOUD) eserevas wea renee ee. 48 Microdispodides pholidotus...............|.......00.0... 48 WMC OMUS DUIS ODOUDTUS cr) Gare) ites ene eaeEmusonlinrne yee i 57 IVUCIILECOGTS USNC OTULUNUS era se, arg nye | asi de en eee 57 Gliphidomuastan mettenmeyentena\ ety eae enn een 53 ACR DEHUDESNOTIULULOCCDIVALG), setae ini neere ees arb lRen pened wert enearee 43 TAC UOIOCOMORCOLILISLES Ram reer tai y itera Yer rata 57 Parapygmephorus halictinis.............0.)e 00 cece eee 48 A\CUUDYOSIA? MUP UMS 5% sha osc hbo dob Seon ak cease eease 43 IVICODUGMLCDRGTUUSES DI eit an ee ee 0 38 §2 THE UNIVERSITY SCIENCE BULLETIN bers are characterized by five-segmented legs I (except in Resina- carus), three pairs of propodosomal dorsals, trochanter I with four setae, of which seta c is unmodified, and by coxa I articulating at a right angle to the longitudinal axis of the body. In addition, its presumably most primitive members may possess two claws on tarsus I, have coxa IV similar in shape to that of the remaining legs, have a pair of setae on the opisthosomal venter, and possess (at least in Dolichocybe and Pyemotes) males having a well-developed gnathosoma, functional mouthparts, and relatively unmodified sex- ual structures and hind legs. The subfamily Acarophenacinae consists of a unique group whose members possess only four of the specialized characters listed in Tables III and IV. It shows closest affinities to primitive pyemotine genera in that the adjacent margins of the propodosoma and hysterosoma are contiguous or nearly so, in having a pair of opistho- somal ventral setae (except in Caraboacarus), in that coxa IV re- sembles that of the remaining legs, in the shape and extent of the ventral apodemes, in that the palps may be fused and the chelicerae large, and in the presence of the gnathosoma and rela- tive lack of modification of sexual structures in the male. Except for the genus Caraboacarus, the subfamily forms a morphologi- cally distinct group based on the loss of ventral setae, the loss of pseudostigmata, the loss of claws, and in having the male genital opening situated ventrally. Caraboacarus alone resembles the more primitive pyemotines in that leg I is five-segmented, in having pseudostigmata, in having the gnathosoma free and not recessed into the propodosoma, and in having the propodosoma entire rather than bipartite. More specialized Acarophenacines resemble the higher pyemotines in having legs I with four segments, trochanter I with three setae, claw I sometimes greatly enlarged and sessile, tarsus I sometimes with a mediolateral sensory spot, and legs I arising oblique to the longitudinal axis of the body. These char- acters are thought to have been independently derived. All species of acarophenacines whose habits are known are parasites of in- sects or their eggs. In the reduction of the ventral setae but retention of the opistho- somal ventrals, in the structure of the ventral apodemes, shape and structure of the propodosoma, position of the stigmata, general size and shape, and in the retention of a functional gnathosoma in the male, members of the Acarophenacinae resemble, and are thought to be closely related to, members of the family Tarsonemidae. The THe FAMILY PYEMOTIDAE 83 origin of the highly specialized Podapolipodidae is less clear but is probably closest to the tarsonemid-acarophenacine groups. No seemingly primitive members of the Podapolipodidae were avail- able for study and observations of specialized forms, e. g., Podapoli- pus, give little information as to their derivation. Despite the apparent difference in the number of derived char- acters between the two groups (due to the fact that some, but not all pyemotines possess derived characters found throughout the Pygmephorinae; see Table V), the distinction between the Pyemo- tinae and Pygmephorinae is not well defined, a gradual transition occurring between the two groups. The point of division between them is therefore arbitrary and is based upon few characters. Weighting of characters (number of segments of leg I, conforma- tion of seta c of trochanter I) has therefore been done deliberately in this case with an eye toward simplification of the classificatory structure. The relationships of the Pygmephorinae have been obscured by earlier authors, who assigned all known species except the two belonging to the hitherto monotypic genera Microdispodides and Microdispus to the single genus Pygmephorus. Because of the close relationship between the two groups, primitive pygmephorines of the tribe Pygmephorini (the latter containing 11 derived char- acters ) share the following characteristics with at least some spe- cialized members (e. g., Siteroptes) of the subfamily Pyemotinae: extreme physogastry, three pairs of propodosomal dorsals, vestiges of the intersegmental line between tibia and tarsus I, four setae on trochanter I, at least apodemes III and IV complete and well developed, apodeme V often present in addition, and leg I articu- lating more or less at a right angle to the longitudinal axis of the body. All of these characteristics are lost within various members of the tribe, but the last-named is found again, perhaps acquired secondarily, in certain members of the genus Pseudopygmephorus of the Neopygmephorini. The tribes Microdispini and Neopygmephorini are considered almost equally specialized, their origins obscure but near the Pygmephorini. The former contains 11 derived characters, the latter 14. Microdispines are characterized in that seta c of trochanter I is setiform and in having lost, except in some Premicrodispus, the second pair of propodosomal dorsals as well as a pair of setae from hysterosomal segment V. In most forms (exception, Peponocara) $4 THe UNIVERSITY SCIENCE BULLETIN the gnathosoma is elongate and often acuminate, the peritremes are widely separated and troughlike or guttate and the first hysterosomal segment is well developed, usually covering at least the posterior half of the propodosoma. As indicated by general body conforma- tion, the articulations of coxae I, the shape and placement of the ventral apodemes, and the tripartite posterior margin of the posterior ventral plate, some Microdispus (Premicrodispus) show many char- acters in common with certain species of Pseudopygmephorus but the relationship between the two genera is obscure. The former is considered herein to be the most primitive from within the Micro- dispini while those forms associated with army ants and culminating with Glyphidomastax and Perperipes are thought to be the most specialized. These latter forms are characterized by the hyper- development of the first hysterosomal tergum, the loss of claws from as well as other modifications of leg IV, the elongation of the palps for piercing, and other specializations. Phenetic relation- ships within the members of the Microdispini are easily established but phylogenetic relationships are obscure. Most members of the Neopygmephorini differ from those of the Pygmephorini, as well as from most of the Microdispini, in pos- sessing two, rather than one or three pairs of propodosomal dorsals. In addition, seta c of trochanter I is bladelike in all forms. Because of the multiple loss and reappearance of many characters, the rela- tionships among the members of this tribe are not clear. Figure 6 illustrates the author’s concept of the relationships within the Tarsonemoidea in accordance with the outline presented above. The horizontal arrangement of categories is a matter of convenience. DESCRIPTION OF THE FAMILY PYEMOTIDAE FEMALES Description. Small mites, 85 to 400. long, of varying shapes; color pearly-white translucent to brown; chelicerae styliform, en- larged bladelike, or (rarely) vestigal; peritremes usually dorsal and of various shapes, rarely (some Microdispini) in “shoulders” and directed anteriorly, or absent (only in Pavania and Dolichocybe ); propodosomal dorsum entire or bipartite, bearing one to three pairs of setae; pseudostigmata usually present; anterior ventral plate with two to six pairs of setae; with four visible dorsal segments bearing, respectively, two, one, one or two, and two pairs of setae; fifth segment usually present, commonly enclosed by fourth, bearing zero to three pairs of setae; first tergum overlapped by, contiguous THE FAMILY PYEMOTIDAE 85 with, or overlapping propodosoma, latter condition present to vary- ing degrees but only rarely (Perperipes, Glyphidomastax ) with free margins which cover the gnathosoma and basal segments of legs I and II; posterior ventral plate with three to six pairs of setae, opisthosomal venter with one pair or none; four pairs of legs always present; legs I four or five segmented, bearing zero to two claws, pulvillus of this leg usually lacking (present only in Pavania and Dolichocybe fusiformis); seta c of trochanter I often modified; leg IV with five segments (tibia and tarsus fused but distinct in Caraboacarus ) and usually with claws; coxa III usually shorter than coxa IV; trochanter IV invariably much longer than femurogenu IV. Present knowledge concerning males is too incomplete to merit a formal description; information concerning them can be found on pp. 68-73 (morphology), in the key to the genera of the sub- family Pyemotinae, and in the various descriptions. Key to the Families of the Tarsonemoidea (Females ) l(a). Hind coxa and trochanter short, each much less than half as long as the greatly elongate and cylindrical femurogenu; hind tarsus without claws, terminating in two elongate, whiplike setae; coxa III elongate, always ab least twiceras) long astcoxa LVS. 40502525. Tarsonemidae 1(6). Hind coxa and trochanter longer, each always much more than half as long as femurogenu; hind tarsus present or not and with or without claws; coxa III never more than one and one-half times as long as coxa 2(a). Usually with three pairs of legs or less; if four pairs are present, then the gnathosoma is free and distinct and the propodosoma is not bipar- tite pseudostigmata lacking. ss: 2 2b een: Podapolipodidae 2(b). With four pairs of legs; pseudostigmata present or, if lacking, then gna- thosoma partly or completely enclosed within propodosoma......... 3 3(a). First hysterosomal tergum greatly expanded anteriorly and laterally so that its free, striate margins project over the gnathosoma and basal segments of legs I and II; segment V with three pairs of setae; seta c of trochanter I modified or absent; usually brown, circular to broadly el- lipticalanvdorsoventralvaspect.: .1- oe ceo nee Scutacaridae 3(b). First hysterosomal tergum of varying shapes and sizes, but as described above only in two genera (Perperipes, Glyphidomastax) associated with doryline ants; if margins enlarged, free, and striate, then seta c of tro- chanter I setiform, segment V with two pairs of setae, gnathosoma greatly elongate, palps sharply acuminate...............Pyemotidae Key to the Subfamilies of Pyemotidae (Females ) l(a). With three pairs of anterior ventral setae or less; pseudostigmata often absent; chelicerae enlarged, distinct; opisthosoma short, usually not longer than distance between anterior margin of coxa III and posterior margin of coxa IV; mites oval or rhomboidal in dorsoventral aspect... . Acarophenacinae, p. 129 1(6). With at least four pairs of anterior ventral setae; pseudostigmata pres- ent; chelicerae usually small, indistinct; opisthosoma usually distinctly longer than distance between anterior margin of coxa III and posterior mere of coxa IV; mites variously shaped but only rarely oval or rhom- (OUCOTE emso ne MIRA cain adh ae aan aps abel Ie Lig pat, ct ICU RR eS MI agi deed 2 86 THE UNIVERSITY SCIENCE BULLETIN 2(a). Leg I usually with five segments, if four then coxa IV triangular (Res7- nacarus); coxa I articulated to anterior ventral plate at a right angle to longitudinal axis of body or nearly so; seta c of trochanter I seti- TKO) 10 Cee Ete te mae ales re eee Leia an re a Inher Tots Pyemotinae, p. 86 2(b). Leg I with four segments; coxa IV quadrate or rectangulate; coxa I articulated as above or distinctly obliquely; seta c of trochanter I seti- fOFMPORANOL Ree OE OL Laan ene re eran Pygmephorinae, p. 147 Subfamily PyYEMOTINAE DEsCRIPTION OF NON-GRAVID FEMALE. Mites of small to moderate size, 150 to 400y. long; shape, elongate-elliptical, elliptical, or fusi- form; posterior margin of hysterosoma evenly rounded, broadly rounded, or truncate; integument weakly to moderately sclerotized, lacking striae, pebblings, or scalelike engravings. Gnathosoma. Always free, normally directed anteriorly or rarely anteroventrally, never largely or completely hidden within propodo- soma; palps fused to gnathosoma or free and moveable; chelicerae enlarged and distinct or small and indistinct. Propodosoma. Dorsum. Always entire, pseudostigmata present; three pairs dorsal propodosomal setae; posterior margin of propodo- soma only meeting, barely overlapping, or barely overlapped by anterior margin of first hysterosomal tergum; esophageal ‘finned” structures present or absent. Venter. Anterior ventral plate with four to six pairs of setae. Hysterosoma. Four to five visible segments, segment V com- pletely enclosed dorsally and posteroventrally by tergum IV or not; segment V with one to three pairs of setae; posterior portion of vaginal wall sclerotized or enclosed in enlarged, elliptical to globose pumplike structure or not; at least one genital sclerite present, its shape variable; opisthosoma not distinctly shorter than distance be- tween anterior margin of coxa III and posterior margin of coxa IV, or if so, then body usually elongate; opisthosomal venter with or without single pair of setae. Legs. Leg I four (rarely) or five-segmented, if four-segmented then coxa IV triangulate; coxa I, as seen from beneath, articulated at a right angle to the longitudinal axis of the body or nearly so; claws present on legs I to IV; trochanter I with four to five setae, seta c setiform; pinnacula absent; coxa IV triangular, quadrate, or rectangulate. THe Famity PYEMOTIDAE 87 Key to the Genera of the Subfamily Pyemotinae Tk(GR) STN Ley Katee Soa Sine Se cis a Ree ek 8 as Sere RT aene OU RcgeiN NY La ae Baek ca nano 7 H1k(G) Ree rane] CSR eee cea icee acy caer et eee ao a ott f0 a ee ine CE eee ER 2 2(a). Leg I with two claws; peritremes lacking....................... 6 2(b). Leg I with one claw; peritremes present...................---.-- 3 3(a). Leg I with four segments, tibia and tarsus fused.......Rest¢nacarus, p. 128 3(6). Leg I with five segments, tibia and tarsus separate............... 4 4(a). Coxa IV as coxa III, triangular; palps fused to gnathosoma (fig. 21) Pyemotes, p. 93 4(b). Coxa IV rectangular or quadrate; palps free..................... 5 5(a). Posterior ventral plate tripartite; claw I sessile (fig. 29)............ Trochometridium, p. 116 5(6). Posterior ventral plate entire; claw I distinctly pedicellate.......... Siteroptes, p. 101 6(a). Palps arising ventrolaterally from gnathosoma; chelicerae large, fal- CATCH UM eae ope elena ibe itin ya cn ce atcne Jeirsh tema riintan 2 rac GH ee csc Pavan, p. 87 6(6). Palps arising laterally from the gnathosoma; chelicerae small, indis- UO VERS Wess On eee eater ttre at A nL nica tiers nu ca ReeS te ae Dolichocybe, p. 89 7(a). Tibia III with greatly enlarged spinelike seta which reaches tip of tar- SUSZOTMEATIVESO Meine Mec nie Meee nats aco ae Nei NS ack SESE iesinacarus (D) palbiovaelelackimnessuchwarsetaynanasst creas ei se ick eee 8 8(a). Gnathosoma well developed, or if reduced, functional (figs. 15, 23). . .9 8(6). Gnathosoma reduced to a lobe, mouthparts lacking (figs. 27, 99)... 10 9(a). Leg I with two claws; leg IV normal (fig. 15)........... Dolichocybe 9(b). Leg I with a single claw; leg IV modified, tonglike........Pyemotes 10(a). Leg I with single pedicellate claw; propodosomal dorsum with four DAITSHONISC LACH a ctem acs ere ales een ARE UT net ai ae Ohne al Seen ee Siteroptes 10(6). Leg I with only broad, flaplike pulvillus; propodosomal dorsum with threecpains Olsetae sit sn eat yo cee le ea Trochometridium Pavania Lombardini (Figs. 17-18) Pavania Lombardini, 1949, Redia (Firenze), 34: 67-74. Type, Pavania fusi- formis Lombardini, 1949, by original designation. This monotypic genus is very closely related to Dolichocybe Krantz, resembling the latter in that the peritremes are lacking or at least not visible, in that the opisthosomal plate is bipartite, each portion bearing an indentation or suckerlike structure, in the struc- ture and conformation of the terminal opisthosomal segments, in the conformation of the coxae, in lacking coxal setae, and in posess- ing two claws on leg I. It differs from Dolichocybe in various gnathosomal characteristics, chiefly in the structure and size of the chelicerae and in the position and size of the palps. Members of both genera are considered to be among the most primitive of known pyemotids. Only the non-gravid females are known. 88 THe UNIVERSITY SCIENCE BULLETIN Diacnosis. Females separable from those of all other genera except Dolichocybe in possessing two claws on leg I and because they lack visible peritremes. They differ from those of Dolichocybe in that the palps arise ventrally rather than laterally and because the chelicerae are greatly enlarged and falcate. DerscripTION OF NON-GRAVID FEMALE. Length, approximately 120n; width, approximately 68»; distance between anterior sterno- coxal condyles of coxae III, approximately 47p; body moderately sclerotized, coarsely punctate. Gnathosoma. Elliptical in dorsal aspect, only slightly longer than wide, directed anteriorly; palp subconical, movable, arising from the ventral surface, bearing distinct apical tooth and two small ventral solenidia; chelicera greatly enlarged, falcate; ventral pharyn- geal-esophageal pump absent, a series of beadlike esophageal struc- tures present. Propodosoma. Dorsum. Inversely U-shaped, expanding laterally at points of articulation of legs I and II; foramen opening anteriorly; posterior margin meeting but not projecting over or covered by first hysterosomal tergum; pseudostigmata present, located in an- terior one-third, external pseudostigmatal socket subquadrate in shape; posterior pseudostigmatal setae at least twice as long as either of two remaining pairs; peritremes not visible, probably ab- sent; distinct esophageal “finned” structures lacking in propodo- soma. Venter. Anterior ventral plate with six pairs of setae, ventrites I and II each with three pairs; external ventral I not forked; apodemes II increasingly distinct at the sides, bending posteriorly near the midline, incomplete; anterior median apodeme complete, distinct; each posterior marginal apodeme indistinctly arcuate; band of pleated conjunctiva present ventrally between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment entire, meeting but not overlapping propodosomal dorsum; fourth segment enclosing fifth and closely associated with it, bearing two pairs of setae, one pair very long; fifth segment reduced, ventral, bearing large, slitlike vulva, the latter flanked by one pair of short setae. Venter. Posterior ventral plate elongate, entire, well defined anteriorly, bearing six pairs of setae; hind margin of posterior ventral plate entire, broadly rounded; second axillary arising from apodeme IV immediately mesad of anterior sternocoxal condyle IV; external presternals well behind and slightly laterad of internal presternals; THe FAMILY PYEMOTIDAE 89 external poststernals close behind but well mesad of posterior mar- gin of fourth coxal foramina; apodemes III distinct, procurved, not reaching midline; apodemes IV strongest laterally, weakly sigmoid, directed anteromesally; apodemes V lacking; posterior median apo- deme weak, barely furcate posteriorly, terminating on level with internal presternals; opisthosoma (dorsally between segments III and IV, ventrally behind the posterior ventral plate and behind the opisthosomal venters) with bands of pleated conjunctiva; opistho- somal venter divided, forming two lateral “ears,” each bearing a single seta and an oval, granulose disc; an antevulvar plate present in front of segment V, bearing two pairs of tiny setae and an internal sclerite; opisthosoma distinctly shorter than distance between an- terior margin of coxa III and posterior margin of coxa IV. Legs. Rather stout; leg I slightly longer and distinctly thicker than leg II, five-segmented; tarsus I rectangulate, narrower than femurogenu I; lateral sensory spot absent; claws I paired, sessile, moderately developed, not dilate basally, a short, flaplike pulvillus present between them; trochanter I with four setae; trochanter II with two setae, femurogenu II with one; trochanter III with one seta, femurogenu III with one; tarsi II and III each with six tactiles; solenidium of Tibia III absent; claws II-[V moderately stout, sim- ple, subequal in size; pulvilli I-IV well developed, longer than claws; coxa III and IV similar in size and shape, elongate-triangular when viewed in inner ventral aspect; tarsus IV similar in size and shape to tarsus III; ratio of femurogenu IV to tarsus IV approxi- mately | to 1.7. Hasirs. Unknown. Known only from Scarabaeus sacer L. (Co- leoptera, Scarabaeidae). DisTRIBUTION. Known only from the type locality. SPECIMENS EXAMINED. P. fusiformis, Tortona, Italy, June, 1947, on Scarabaeus sacer L. (Syntypes). Dolichocybe Krantz (Figs. 13-16) Dolichocybe Krantz, 1957, Ann. Ent. Soc. Amer., 50: 259-264. Type, Doli- chocybe keiferi Krantz, 1957, by original designation. As mentioned previously (p. 87), this genus shows a close re- lationship to Pavania Lombardini. It is retained herein largely be- cause of the structure of the chelicerae and palps. The author has not seen the form described as P. fusiformis by Krezal (1959b, p. 565) but comparison of its description with syn- 90 THe UNIVERSITY SCIENCE BULLETIN types of fusiformis indicates that it is a new species which is here tentatively placed as a second species of Dolichocybe until further work clarifies the status of these two genera. In addition to the females of both species, gravid females and males of D. keiferi are known. Dracnosis. Both sexes are immediately separable from all other members of the family except Pavania because tarsus I bears two claws. Females differ from those of Pavania in that the palps arise laterally from the gnathosoma and because the chelicerae are small and needlelike rather than enlarged and falcate. DesCRIPTION OF NON-GRAVID FEMALE. Length, 140 to 190; width, 40 to 65y; distance between anterior sternocoxal condyles of coxa III, 28 to 32u.; hysterosoma elongate-elliptical to subcylindrical; body very weakly sclerotized, without coarse punctae. Gnathosoma. Ovate or elongate-rectangulate, normally directed anteriorly; palp palmate or elongate, if as latter, then cylindrical, bearing four to five short tactile and sensory setae, the most distinct of which is apical and falcate; palp movable but articulations in- distinct; chelicera minute, indistinct; large, irregularly elliptical cir- cumesophageal mass indistinctly present, this not distinctly ventral or bulbose as described for Pyemotes; esophagus narrow, straight, often indistinct. Propodosoma. Dorsum. Rounded or oval, expanding laterally at points of articulation of legs I and I, anterior margin forming cylindrical “cervix” which encloses only gnathosomal base; foramen anterior, foraminal collar absent; posterior margin meeting but not prejecting over or covered by first hysterosomal tergum; pseudo- stigmata present, located in anterior one-third; posterior pseudo- stigmatal setae at least twice as long as either of other pairs; peritremes not visible, probably absent; esophageal “finned” struc- tures absent. Venter. Anterior ventral plate with six pairs of setae, ventrites I and II each with three pairs; external ventral I not forked; apodemes II weak, often indistinct, gently arcuate pos- teriorly, incomplete medially; median apodeme appearing excavated groovelike; posterior marginal apodemes indistinct; large band of pleated conjunctiva present ventrally between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment entire, meeting but not overlapping propodosomal dorsum; fourth segment enclosing apex THE Famity PYEMOTIDAE 91 of hysterosoma, bearing two pairs of setae, one pair very long; fifth segment closely associated with fourth, reduced, ventral, bearing vulva, the latter flanked by a pair of short setae; antevulvar plate present ventrally in front of segment V, bearing two pairs of tiny setae and an internal sclerite. Venter. Posterior ventral plate elongate, poorly defined anteriorly; median trigonal portion of pos- terior ventral plate lacking but latter indistinctly divided into three sections, (fig. 13 and Krezal, 1959b. p. 566) the two anterior each surrounding one of coxae III, the posterior enclosing both of coxae IV (the margins of these sections do not correspond to the ventral apodemes, which are separately distinct); posterior ventral plate with six pairs of setae, its hind margin produced mesally into a flap- like lobe; second axillary arising from apodeme IV immediately be- fore the anterior sternocoxal condyle of leg IV; external presternals well behind and slightly laterad of internal presternals; external poststernals close behind but well mesad of posterior margin of fourth coxal foramina; apodemes III short, recurved posteriorly, not reaching midline; apodemes IV equally short, directed nearly mesad; apodemes V lacking; posterior median apodeme absent; opisthosoma (dorsally between segments III and IV, ventrally behind the opistho- somal venter) with a wide band of distinctly pleated conjunctiva; opisthosomal venter divided, forming two lateral “ears” which join but do not fuse with median lobe of posterior ventral plate, each piece bearing a single seta and an anteromedian depression; opistho- soma shorter than or as long as distance between anterior margin of coxa III and posterior margin of coxa IV. Legs. Short and slender; leg I slightly longer and distinctly thicker than leg II, five-segmented; tarsus I rectangulate, never wider than femurogenu I; lateral sensory spot absent; claws I paired, sessile, moderately developed, not dilated basally, pulvillus present or lacking; trochanter I with four setae, seta c long, seti- form; trochanter II with two setae, fermurogenu II with one; trochanter III with one seta, femurogenu III with one; tarsi II and Ill each with five tactiles; solenidia of tibiae If and III absent; claws II to IV simple, subequal in size; pulvilli II to IV well de- veloped, longer than claws; coxa III elongate-triangular when viewed in inner ventral aspect; coxa IV similar in shape to coxa III but slightly larger; tarsus IV short, rather abruptly constricted basad of pretarsus; ratio of femurogenu IV to tarsus IV approximately Go) Ae 92 THE UNIVERSITY SCIENCE BULLETIN DEscrIPTION OF GRAVID FEMALE (based only on D. keiferi Krantz). Gnathosoma and propodosoma as described for non-gravid female but entire hysterosoma becoming physogastric, globular; diameter of hysterosoma of slide-mounted specimens containing eggs, 245 to 400u.. DeEscRIPTION OF MALE (based only on D. keiferi Krantz). Length, approximately 105 to 115y; width, approximately 60 to 65u.; hystero- soma ovate; conjunctival areas not granulose. Gnathosoma. Subglobose, well developed, apparently functional, chelicerae indistinct but palp and internal pumplike structure at least partly distinct. Propodosoma. Dorsum. Similar to that of female but much broader, bearing four pairs of setae in approximately same location to each other as in female. Venter. Similar to that of female. Hysterosoma. Dorsum. Four visible dorsal segments, the first bearing two pairs of setae, the second one, the third two, the fourth three; fifth segment ventral. Venter. Posterior ventral plate similar to that of female but lacking median posterior lobe; hysterosomal venter apparently fused with segment V, the combined sclerites bearing four pairs of setae, acuminate posteriorly, aedeagus ap- parently opening at apex, aedeagal armature indistinct, perhaps lacking; well defined genital capsule or specialized clasping struc- tures lacking; internal triangulate structure (vesicula seminalis? ) prominent between aedeagal duct and vas deferens. Legs. Unmodified, generally shorter but otherwise similar to those of female. Habits. The only published information concerning the habits of Dolichocybe is found in the description of D. keiferi. Krantz (1957) states that the females give birth to active adult males and females. Gravid females were found under the bark of Acer palmatum Thunb. “at approximately the cambium.” The limbs of the tree from which the type series was taken were dying but the relationship of the mites, if any, to this condition was not known. A series of gravid females taken from the “tunnel of a wood borer” have been seen by this author. They indicate a more probable parasitic method of existence. The form designated as P. fusiformis by Krezal is re- corded by that author from Scarabaeus sacer L. and Tatropium gabrieli. Because of the relatively unmodified posterior portions of the male body, it is probable that at least males of D. keiferi do not THe Famity PYEMOTIDAE 93 grasp and carry the females about prior to and during copulation as is the habit of the higher pyemotids. Distrisution. D. keiferi: Known from Arizona, California, Washington, and Puerto Rico. “D. fusiformis” of Krezal: Known from Italy and France. SPECIMENS EXAMINED. D. keiferi, from Arizona, Washington, California (Syntypes), and Puerto Rico, taken from the following hosts or situations: Acer palmatum, custard apple, from beneath apple bark, and in tunnel of wood borer in red bud. Forms thought to belong to Dolichocybe Krantz: keiferi, Krantz, 1957 Pawvania fusiformis, Krcezal, 1959, not P. fusiformis Lombardini, 1949 Pyemotes Amerling (Figs, 21-24) Heteropus Newport, 1850, Proc. Linnean Soc. London, 2:70 (nom. praeocc. in Insecta, Palisot-Beauvois, 1805). Type, Heteropus ventricosus Newport, 1850, by original designation. Pyemotes Amerling, 1862, Ceska Spole¢nost Nauk, Prague, 1862, p. 54. Type, Piemotes Eccoptogasteri pruni Amerling, 1862, nomen nudum = Heteropus ventricosus Newport, 1850, by designation of Oudemans, 1937. Lais DeFilippi, 1862, Archivio per la Zoologia, ’Anatomia e la Fisiologia, 1(2): 211-215, Tav. XIV (nom. praeocc. in Tunicata, Gistel, 1848). Type, Lais heterogyne DeFilippi, 1862, by original designation. Physogaster Lichtenstein, 1868, Bulletin Entomologique. Seances de la Société Entomologique de la France, (4)8: LXXXVI (nom. praeocc. in Insecta, Lacordaire, 1830). Type, Physogaster larvarum Lichtenstein, 1868, by original designation. larvarum = nomen nudum Oudemans, 1936. Pediculoides Targioni-Tozzetti, 1878, Annali dell’Agricoltura, 1:241. Type, Acarus tritici LaGréze-Fossot and Montané, 1851, Rec. Agronomique de la Société des Sciences, Agriculture, et Belles Lettres, 32(2), by original designation, not Pediculoides tritici Targioni-Tozzetti, 1878. Crithoptes Geber, 1879, Wiener Medizinische Presse, 20(43): 1429. Type Crithoptes monunguiculosis Geber, 1879, by original designation. Sphaerogyna Laboulbéne and Megnin, 1885, Journal de l’Anatomie et de la Physiologie Normales et Pathologiques de 1Homme et des Animaux, 21:14. Nom. nov. for Heteropus Newport, 1850. Physacarus Trybom, 1893, Entomologisk Tidskrift, 14: 121-126. Nom. nov. for Heteropus Newport, 1850. Phthiroides Oudemans, 1936, Festschrift fiir Embrik Strand, 1: 403. Type, Phthiroides megnini Oudemans, 1936, nom. nov. for Sphaerogyna ventri- cosa (Newport) Laboulbéne and Megnin, 1885, not Heteropus ventricosus Newport, 1850. Members of this genus are among the best known of all Acarina. At least one species is a parasite of larval Lepidoptera in various stored grains, in which it may form large populations, these swarm- ing to attack handlers and carriers of the infested material. The 5—1367 94 Tue UNIvERSItTY SCIENCE BULLETIN bite causes mild to severe cutaneous reactions, always accompanied by severe itching and sometimes by nausea and other symptoms. The erruption resulting from a number of bites has been confused with scabies, chicken pox, and other diseases. Severely infested grain has been reported fatal when fed to chickens (see Schamberg, 1910). This paper will not deal with the medically important as- pects of P. ventricosus and related forms since they have been well reviewed by several authors. A nearly complete bibliography of the literature extant in this area is available from the author upon re- quest. Attempts to utilize a species of Pyemotes in the biological control of the boll weevil were unsuccessful (Hunter and Hinds, 1905). Although the author follows Oudemans (1936), in placing at least four forms in the genus, it seems improbable that host speci- ficity is as restricted as believed by that author. Lichtenstein [ 1868, (in Webster, 1883) ], and Herfs (1926), recorded a Pyemotes attack- ing larval Hymenoptera, Coleoptera, and Lepidoptera indiscrim- inately, and the latter author deliberately cultured the parasite suc- cessfully on various hosts. (For detailed listings of hosts, see Herfs, 1926.) More recently Krezal (1959a), describes P. boylei, men- tioning that it is an efficient parasite of both the anthribid Araecerus levipennis Jordan, and of the termite Crypiotermes brevipennis Walker. Because of the medical importance of the genus numerous pub- lications have appeared which are concerned with its bionomics. Probably the most notable are the works of Brucker (1901), Herfs (1926), Monterosso (1934), and Krezal (1959b). Patau (1936), has made detailed cytogenetic studies of one species. Diacnosis. Females may be separated from all others in the family except Trochometridium because the posterior ventral plate is tripartite, its median section elongate and triangular. They are distinguished from Trochometridium in that coxa IV is triangular in the manner of coxa III, because the palps are fused to the gnatho- soma, and by the presence of a pair of opisthosomal ventral setae. DESCRIPTION OF NON-GRAVID FEMALE. Length, approximately 200 to 325y; width, approximately 70 to 125y; hysterosoma elongate, cylindrical to fusiform or rarely rhomboidal; body weakly to lightly sclerotized and without coarse punctae. Gnathosoma. Subglobose to ovate, usually large, normally di- rected anteriorly; palps fixed, fused to gnathosoma; anterior mar- THe Famity PYEMOTIDAE 95 gins of gnathosoma protruding slightly, truncate or gently rounded, anterior extremity of ventral margin marked by second palpal setae; chelicera stout, distinct, aciculate; esophageal pharyngeal pump prominent; dorsobasal “tentorium” (“Chitinspangen” of Krezal) also well defined. Propodosoma. Dorsum. Trapezoidal to sagittate, lateral margin sinuate, anterior margin often projecting over gnathosoma, indis- tinct; foramen opening anteriorly, foraminal collar absent; posterior margin projecting slightly to distinctly over the anterior margin of first hysterosomal tergum; pseudostigmata present, located in middle third; three pairs of propodosomal setae, posterior pseudo- stigmatal pair at least twice as long as either of others; peritremes dorsal, divergent, their elliptical openings near laterobasal angles of gnathosoma; esophageal “finned” structures absent. Venter. An- terior ventral plate with four pairs of setae, (Oudemans, 1936, probably erroneously, describes P. fortuitus as having but two pairs ) ventrites I and II each with two pairs; ventral externals I not forked; coxa I attached at a right angle to longitudinal axis of body or slightly obliquely; apodemes II straight or weakly arcuate, com- plete or incomplete medially, their mesal terminations never sud- denly bending posteriorly; anterior median apodeme distinct, com- plete; posterior marginal apodemes absent or indistinct except laterally; distinct band of conjunctival pleats absent between propodosoma and hyterosoma. Hysterosoma. Dorsum. First segment entire, its anterior margin slightly to distinctly hidden beneath hind margin of propodosoma; cuticular attachments of dorsoventral muscles distinct anterior to dorsal setae II and III; fourth segment bearing two pairs setae, partially enclosing apex of hysterosoma; fifth segment reduced, normally visible only ventrally, bearing a single pair of ventral setae and the apical vulva. Venter. Posterior ventral plate elon- gate, poorly defined anteriorly, tripartite, possessing a large, distinct, median, trigonal area; posterior ventral plate with five pairs of setae; first axillary on or slightly posterior to apodeme III, always mesad of coxa III, second axillary well mesad of anterior sternocoxal condyle IV; external presternal well behind and usually only slightly lateral to internal presternal; one pair of poststernals, these arising from medial trigonal sclerite approximately on line drawn between posterior margins of coxae IV; apodemes III and IV short, the former gently arcuate, the latter gently sigmoid, neither reaching 96 THe UNIVERSITY SCIENCE BULLETIN trigonal area; median apodeme absent; opisthosoma with wide band of distinctly pleated conjunctiva dorsally between segments III and IV, ventrally between the opisthosomal venter and segment V; opisthosomal venter entire, bearing a single pair of setae. Legs. Usually slender, moderately elongate; leg I about as long as and distinctly thicker than leg I, five segmented; tarsus I quadrate to subconical, never wider than femurogenu I; lateral sensory spot absent; claw I arising from a short, stout pedicel, often appearing sessile or nearly so; thumb prominent to absent; tro- chanter I with four setae, seta c long-setiform; trochanter II with three setae, femurogenu II with three; trochanter HI with two setae, femurogenu III with three; tarsi II and III each with five to seven tactile setae, solenidium of II near middle of segment, solen- idia of tibiae II and III present or absent; claws II to IV simple or with weak basal spur, subequal in size; pulvilli II to IV well de- veloped, longer than claws, coxa III elongate-triangular when viewed in inner ventral aspect; coxa IV similar in shape to coxa III but shorter and stouter; tarsus IV short, not or slightly longer than tarsus III, constricted abruptly basad of pretarsus; ratio of femuro- genu IV to tarsus IV, approximately 1:2.0 to 1:2.5. DescripTION OF Gravip FEMALE. Externally as in non-gravid female but opisthosomal conjunctival bands unpleating to form turgid, milky globe which may reach 2 mm. in diameter; segments IV and V visible and adjacent at apex of globe. DeEscripTION OF MALE (based only on males of one species in col- lection of author plus description of P. boylei Krezal, 1959a). Length 190 to 200»; width, approximately 100 to 120u; hysterosoma ovate; conjunctival areas not granulose. Gnathosoma. Rounded, well developed, functional; palpi fused to gnathosoma as in female; chelicerae reduced but distinct, needle- like; esophageal-pharyngeal pump well defined. Propodosoma. Dorsum. Large, trapezoidal or with sides sinuate, bearing three or four pairs setae, if four then fourth pair tiny, slightly anterior to and well mesad of anteriormost of remaining setae. Venter. Similar to that of female but shorter. Hysterosoma. Dorsum. Dorsal segmentation modified, first plate subcircular, bearing three pairs setae, second reaching genital cap- sule (“Kopulationsapparat” of Krezal, 1959b), bearing but a single pair; posteriormost portion of body constricted to form capsule, this concave posteriorly, bearing single short pair of dorsal setae. i THE Famity PYEMOTIDAE 97 Venter. Apodemes III, IV, and V well developed, posterior median apodeme absent; setation of posterior ventral plate similar to that of female; opisthosomal venter not visible as definitive sclerite; one or two pairs of setae ventrally on genital globe; genital capsule with one dorsal pair and one ventral pair of suckerlike structures; aedeagus tubular, well sclerotized. Legs. All legs five-segmented; legs I and II similar to those of female but shorter; leg III distinctly longer than I, II, or IV, coxa IV subquadrate, trochanter IV distinctly arcuate, width uniform, apex rounded. Habits. These mites were generally believed by early workers to feed upon stored grain and their true nature as parasites of grain- infesting insects was not suggested until 1883 (by Webster). The following information is taken largely from Herfs (1926), and Krezal (1959b ), and deals with P. herfsi and P. scolyti respectively. All known members of the genus are external parasites of insects, usually killing the host species. Hosts are recorded from the follow- ing orders: Homoptera, Coleoptera, Lepidoptera, Diptera, and Hymenoptera. In nearly all cases, the hosts are attacked in their immature stages, but there are occasional records of the parasitiza- tion of adults [Callidium variable (Cerambycidae), vide Brucker (1901); Isosoma (= Harmolita) tritici (Fitch) (Eurytomidae ), vide Webster (1910); Tineola biselliella Hummel (Tineidae), vide Herfs (1926) ] and of eggs [ Magicicada septendecim L. (Cicadidae ), vide Marlatt (1898); and Fidia viticida Walsh (Chrysomelidae ) accord- ing to Webster (see Reuter, 1900) ]. Certain larvae may be immune to attack. Herfs mentions that the mites in his cultures failed to successfully parasitize larvae of Ephestia kuhniella Zeller although they readily attacked and killed those of Tineola biselliella and Pec- tinophora (= Gelechia) gossypiella (Saunders). A larval ceram- bycid and the larvae of “Ringelspinner” were likewise not parasi- tized. Herfs concludes that this immunity exists by reason of the depth or hardness of the cuticle. Still other larvae, e. g., those of Diloba caerulocephala, were attacked and killed without enabling the parasite to complete its development. Both males and females pass through all immature stages while still within the enormously distended, globular opisthosoma of the mother mite, and are born as adults. Most males are born during the first half of the birth period and normally are present only in small numbers (three to five percent of the total number of off- spring at 25° C.). They seldom leave the body of the mother and 98 THE UNIVERSITY SCIENCE BULLETIN are parasitic upon it. At 25° C., males of herfsi removed from the maternal “opisthosomal globe” died 16 to 25 hours after removal, presumably from starvation, and if placed on the bodies of normal, healthy hosts did not feed. Males are usually found resting quietly on the opisthosoma of the mother, for the most part near the vulva, or (P. scolyti) the first born male may remain partially within the vagina of the mother, only the posterior portion of its body pro- truding through the vulva. They may (herfsi) or do not (scolyti) assist in the birth of the females by grasping the latter with the hind legs and pulling, shifting the grasp, pulling again, etc. This aid is not necessary since birth of the females continues smoothly in the absence of males although taking slightly longer (average birth time of females of herfsi with males, 171 seconds, average time without males, 217 seconds). In scolyti the combined processes of birth and mating require less than a minute. As soon as (herfsi) or shortly before (scolyti) the newly born female completely leaves the birth canal, the male initiates copu- lation. In herfsi, he grasps the female and adjusts, with the assist- ance of the hind legs, her body until contact is made. This entire process of adjustment and mating may take 10 minutes or more, although actual copulation probably lasts only 20 to 30 seconds. Females of scolyti push past the intruded body of the male until they reach the level of the hind legs, at which point they are grasped and mated. Brucker (1901) observed copulation to take place on the host rather than on the opisthosoma of the mother. This may be a specific difference, since the material investigated by Brucker parasitized Callidium sp. (Curculionidae) while Herfs’ forms were found on lepidopterous larvae. Because of the config- uration and length of the hind legs in the male, Oudemans (1936) designates this first form P. bruckeri, nom. nov. Males of herfsi normally assist only in the birth of the females, paying no attention to emerging males. They likewise ignore mated females, although one or more males may attempt to disrupt pairs in copulo. Early in the birth period of the mother, only a single herfsi male is usually found on the opisthosoma but as this period progresses, several may be present. The second and succeeding males of scolyti likewise take up parasitic residence upon the ma- ternal hysterosoma. They seemingly function as replacements for the male situated partially within the vagina, and wait to assume its task of inseminating each female as she is born. Shortly (within 10 minutes) after being mated, the young female leaves the body of the mother and seeks food. If, however, males THE FAMILY PYEMOTIDAE 99 are lacking on the maternal hysterosoma, young females do not leave but remain resting upon the latter for some time awaiting the birth of a male. Only after a full day or longer do they leave. Unmated females seemingly suffer a high mortality, only about 10 percent of those in experimental cultures surviving to produce off- spring. Such females are arrhenotokous, producing only males. The total number of offspring produced compares favorably with that of mated females of the same size. According to both Krezal and Herfs, young females die of starva- tion within 48 hours of their birth if food is not found. This state- ment makes dispersion and perhaps overwintering difficult to explain since it is obviously the young females which must disperse. Herfs found that young females unable to find a host would sometimes return to the mother and feed in the manner of the male on the parents hysterosoma. Rarely this filial parasitism became perma- nent and ended with the death of the mother and the production of offspring by the daughter. More commonly it was quite transi- tory and resulted in no visible harm to the mother while seemingly providing the daughter with enough food to prevent at least im- mediate starvation. Young females of both herfsi and scolyti are stimulated by moving objects soon after leaving the mother, stretching the body toward the source of the stimulus. Such behavior indicates dispersion by phoresy, a phenomenon which has been observed only rarely in Pyemotes despite its widespread occurrence in other members of the family. Krezal observed young females of scolyti attached to adult Scolytus scolytus to live somewhat longer than isolated females of the same age, from which he concluded that such phoretic females fed from body exudates of the host without obtaining enough nourishment to initiate physogastry. In view of the fact that young females starve so quickly, the questions of their survival between generations of the host and of overwintering remain essentially unanswered. As mentioned by Krezal, perhaps the phoretic period is ephemeral, serving only to bring the females into fortuitous contact with one of a number of possible hosts. This could explain their rare and sporadic occur- rence in the nests of solitary bees and social wasps, hosts in which an overwintering period would be required of the young females. Presumably, such a period would result in their migration or death, thus preserving the host from further attack until the next “acci- dental” infestation occurred. 100 THe UNIVERSITY SCIENCE BULLETIN Upon finding a host, the young female selects a favorable location and inserts her chelicerae, an action which may elicit striking or curling movements from the host. Within a short time, the host begins to show signs of paralysis, this increasing until the larva is completely incapable of movement. Time elapsed between onset of attack and complete paralysis depends upon the number of punctures made, but generally is two to four hours. The same larva may be attacked by several mites, Herfs having counted a maximum of 178 on a single host. The parasite develops quickly, the opisthosoma increasing in both length and width until the mite assumes a form (the “Zitronen- stadium” of Herfs) quite similar to an incandescent light bulb. This stage is usually reached in about 24 hours. Complete develop- ment, in which the opisthosoma becomes a huge, turgid sphere may be completed in another 24 hours, although it usually requires seven to eleven days to reach maximum size. At 25° C. the average time from initial insertion of the chelicerae to birth of the first offspring is 9.5 days in P. herfsi. The size of the opisthosomal globe is well correlated with the number of offspring produced. In cases of multiple parasitism the size of the opisthosoma will be reduced, and females concerned will produce a relatively small number of offspring, sometimes averaging as low as six or seven. Females of herfsi reared singly produced a maximum of 284 offspring and averaged 128 at 25° C. and 150 “at room temperature.” The birth period, usually terminated by the death of the female, ranged from 9 to 33 days and averaged 17.3 days. The average life span of the females in Herfs’ cultures at 25° C. was 26.8 days as measured from the beginning of feeding to the death of the mite. The average life span of the male is much shorter, since it dies with the mother mite. It was found that longevity of herfsi males was also positively correlated with amount of sexual activity. The temperature of the microhabitat was also an important factor affect- ing both longevity and fecundity of both sexes. Several previous workers have found 25 to 30° C. to be optimum and Herfs noted a loss in the ability of the mother to give birth at 35° C. DisrriBuTION. At least P. ventricosus, cosmopolitan. SPECIMENS EXAMINED. At least two species (one near scolyti, one near boylei) from many parts of the United States from the follow- ing hosts or situations: scolytid beetle, nest of wasp, man, cotton crop residue, pink bollworm larvae, scale (?Chrysomphalis) on Quercus macrocarpa Michx. THE FamMiILy PYEMOTIDAE 101 Species thought to belong in Pyemotes: alastoris (Froggatt, 1894) (Heteropus) Nomen dubium anobii Krezal, 1959 attelabrinus (Debey, 1849) (Dermaleichus ) beckeri Krezal, 1959 boylei Krezal, 1959 bruckeri Oudemans, 1937 dryas ( Vitzthum, 1923) ( Pediculoides ) Eccoptogasteri pruni Amerling, 1862 ( Piemotes) Nomen nudum fortuitus (Oudemans, 1936) ( Pediculoides ) hartigi Oudemans, 1937 Nomen dubium herfsi (Oudemans, 1936) ( Pediculoides ) heterogyne (DeFilippi, 1862) ( Lais) larvarum (Lichtenstein, 1868) (Physogaster) Nomen nudum megnini (Oudemans, 1936) ( Phthiroides ) monunguiculosus (Geber, 1879) (Chrithoptes ) schwerdtfegeri Krczal, 1959 scolyti (Oudemans, 1936) ( Pediculoides ) takeuchii (Sasa, 1947) ( Pediculoides ) tritici (LaGreze-Fossot and Montané, 1851) ( Acarus) ventricosus ( Newport, 1850) (Heteropus ) Siteroptes Amerling As defined herein, this genus includes several forms placed by various authors in Pygmephorus and other genera. Although in- cluding a series of forms which range from a relatively primitive Dolichocybe-like species to forms very closely related and ancestral to the mites of the subfamily Pygmephorinae, the author has pre- ferred to include them in a single genus divided into three sub- genera. In addition to these, the author has seen a single damaged specimen of an undescribed species which is obviously quite closely related to primitive forms of Pediculaster, and which, when de- scribed, should undoubtedly form a fourth subgenus. A relatively large number of forms seemingly occur in North America. Fifteen species are thought to be included in the material, nearly entirely North American, studied during the course of this work. Most species are thought to be phytophagous or saprophagous. Only S. aradii Krezal is known to be phoretic upon insects. Diacnosis. The females of this genus may be distinguished from those of all other genera in the family by the following combination of characters: leg I with five segments; single claw of tarsus I simple, pedicellate, lacking thumb; chelicerae small, indistinct; palps free. 102 THe UNIvERSITY SCIENCE BULLETIN DeEscriPpTION OF Non-GrRAvID FEMALE. Length, approximately 155 to 390»; width, approximately 75 to 175y; hysterosoma elongate- cylindrical to broadly elliptical; body usually weakly, occasionally moderately sclerotized, and without coarse punctae. Gnathosoma. Squarish to elliptical, normally directed anteriorly; palp distinct, free, short to moderate in length, its apex usually dentate, one tooth usually much longer than the rest; chelicera tiny, indistinct; esophageal-pharyngeal pump absent; esophagus straight, thick, usually indistinct. Propodosoma. Dorsum. Soft, shape variable according to mount but generally elongate-trapezoidal or elongate with sides broadly sinuate, expanding laterally at points of articulation of legs I and II, anterior margin forming cervix which encloses base of gna- thosoma, this cervix continuing around gnathosoma to form indis- tinct foraminal collar which may or may not become excavated on ventral surface; posterior margin meeting or barely covered by first hysterosomal segment; pseudostigmata present, arising in posterior one-half; three pairs of propodosomal setae, posterior pseudostig- matals always distinctly longer than other two pairs; peritremes dorsal and anterior, or arising anteromesally near stigmatal setae, elongate-cylindrical to guttate or (rarely) circular; esophageal “finned” structures present, indistinct or distinct, usually one short structure in gnathosoma or propodosoma followed by a more or less globular or quadrate structure, this followed by a third structure which may be similar to the preceding one or by a shorter structure whose posterior margin is distinctly defined and recurved cephalad. Venter. Anterior ventral plate with four (rarely) to six pairs setae, ventrites I usually with three pairs; external ventrals I forked or not; coxa I attached at a right angle to longitudinal axis of body; apo- demes II distinct, weakly sigmoid, usually incomplete medially, always bending posteriorly as they approach anterior median ap- odeme; anterior median apodeme distinct, complete, posterior mar- ginal apodemes well developed only laterally; large band of pleated conjunctiva indistinct or absent between propodosoma and hystero- soma. Hysterosoma. Dorsum. First segment entire, meeting or extend- ing slightly over posterior margin of propodosomal dorsum, third and fourth segments with one or two pairs setae; fifth segment usu- ally overlapped by fourth segment, enclosing apex of hysterosoma, bearing three pairs of setae; vulva apical, usually distinct, vagina enclosed apically by sclerotized, U-shaped, globose to elliptical THE FamMiIty PYEMOTIDAE 103 pumplike enlargement or vaginal wall merely sclerotized apically; anterior and median genital sclerites absent; posterior genital scle- rites usually also absent; a globose or omega-shaped vaginal structure sometimes distinct. Venter. Posterior ventral plate poorly defined anteriorly, entire or divided behind coxae III into two parts (fig. 19), the margins of these parts not corresponding to the ventral apo- demes (epimera of authors), which are separately distinct; pos- terior ventral plate with four (rarely) to six pairs of setae, hind margin of posterior ventral plate trilobed, with single median lobe, or entire; external presternals well posterior to and directly behind or only slightly laterad of internal presternals; external poststernals rarely absent, if present distance behind coxae IV variable; apo- demes III variable, present only laterally or complete, joining pos- terior median apodeme with anterior sternocoxal condyle of III; apodemes IV variable, distinct only laterally, or complete or nearly so, always joining anterior sternocoxal condyles of legs IV; apodemes V present or absent; posterior median apodeme variable, distinct only in region of apodemes IV, or well developed; opisthosoma (dorsally between terga III and IV, ventrally between the opis- thosomal venter and segment V) with or without a wide band of distinctly pleated conjunctiva; opisthosomal venter entire, devoid of setae except in some Metasiteroptes. Legs. Variable in length and thickness; leg I at least slightly thicker than leg II, five segmented; tarsus I quadrate, rectangulate, or elliptical, never wider than femurogenu I; lateral sensory spot absent; claw I single, simple, hooklike, moderately developed, dis- tinctly pedicellate, pedicel stalklike, short, but usually distinctly longer than wide; thumb absent; trochanter I with three (rarely) or four setae, seta c long, setiform; trochanter II with three setae, femurogenu II with one to three; trochanter III with one or two setae, femurogenu III with one or two; tarsi II and III each with six (rarely) or seven tactiles; solenidium of tarsus II in proximal half; sensory pegs of tibiae II and III present; claws II to IV simple, sub- equal in size; pulvilli II to IV variable; coxa III elongate-triangular to elongate subcordate in inner ventral aspect; coxa IV triangular and smaller than coxa III (Metasiteroptes), quadrate, or short- rectangulate, usually widest at base; femurogenu IV with one seta or none; tarsus IV usually short and abruptly constricted basad of pretarsus; ratio of femurogenu IV to tarsus IV 1:2.1 to 1:6.2 (only rarely over 3.8—if as large as 6.2 then mite very large, 385y. long and having trochanter I with only three setae). 104 THE UNIVERSITY SCIENCE BULLETIN DEscCRIPTION OF GRAVID FEMALE. Gnathosoma and propodosoma as described for non-gravid female, but entire hysterosoma becom- ing physogastric, globular. Diameter of hysterosoma of females as great as three millimeters. DEscrIPTION OF MALE (based on three species of Siteroptes s. str. and Siteroptoides). Length, 120 to 165; width, 70 to 95y,; hystero- soma generally ovate, sharply constricted in vicinity of legs IV, opisthosoma forming enlarged connate, globose, or subtriangular copulatory structure; conjunctival areas not granulose. Gnathosoma. Vestigal, present only as reduced lobe of varying shapes, this lobe bearing one median dorsal pair, one anterodorsal pair, and one posteroventral pair of tactile setae, in addition to two anteroventral pairs of solenidia. Propodosoma. Dorsum. Much broader than that of female, bear- ing four pairs of setae. Venter. Similar to that of female but shorter and broader; with either five or six pairs of ventral setae. Hysterosoma. Dorsum. Segmentation modified, first plate large, usually rounded posteriorly, bearing three pairs of setae; second plate much smaller, strongly or weakly defined, variable in shape, bearing one or two pairs of setae; third plate forming basal portion of genital capsule, with two pairs of setae, both pairs usually sensory, the anterior pair much the larger; aedeagal capsule large, at- tenuate apically, bearing two pairs of apicoventral setae; aedeagal armature well sclerotized, divided longitudinally into halves, pointed apically, short, aedeagal base not greatly elongate, not plainly re- curved; internal “vesicula seminalis” usually globular, possessing lateral “wings.” Venter.. Apodemes IV and V well developed, apodemes III strongly developed entirely or only laterally; posterior median apodeme present, connecting apodemes V with apodemes IV or often ending anteriorly to join apodemes III; anterior ex- tremities of apodemes V joining posterior median apodeme, char- acteristically obfurcate between coxae IV; setation of posterior ven- tral plate as described for female; hind margin of posterior ventral plate well posterior to coxae IV, enclosing genital capsule; opistho- somal venter apparently fused with genital capsule, not distinct. Legs. All legs five-segmented; length, claws, setation, and pro- portions of legs I-III similar to those of female; leg IV tonglike, modified for clasping, short; coxa IV subquadrate, trochanter IV not distinctly arcuate, tarsus IV greatly reduced, two claws present but at least one greatly deformed, scooplike or irregular. THe Famity PYEMOTIDAE 105 Key to the Subgenera of Siteroptes (Females ) 1(a). Coxa IV triangular; posterior ventral plate divided and appearing as two distinct plates, each bearing a pair of legs Metasiteroptes, new subgenus, p. 105 1(6). Coxa IV quadrate or rectangulate; posterior ventral plate entire... .2 2(a). Peritremes guttate, circular or rarely elongate-guttate, usually not much longer than wide, separated by less than one of their lengths; femurogenu II with three setae; apodemes V always well developed; ventrites II each with two or three setae Siteroptoides, new subgenus, p. 113 2(b). Peritremes usually elongate, at least twice as long as wide, or if circular or short-guttate, then separated by at least one of their own lengths; femurogenu II with two or three setae; apodemes V rarely well de- veloped; ventrites II each with two setae. . . Szteroptes sensu stricto, p. 108 Metasiteroptes, new subgenus (Figs. 19-20) Type, Siteroptes (Metasiteroptes) macer, new species. This subgenus is established to accommodate two species which retain several of the more primitive characters found in Dolicho- cybe, etc. including short, oblique, incomplete apodemes III, and triangular coxae IV. They are easily distinguished from all other known species of Siteroptes and perhaps deserve generic status. Dracnosis. For females, see the key characters. Males and im- matures are unknown. DEscrIpTION OF NON-Gravip FEMALE (based upon S. macer). Hysterosoma elongate-cylindrical; body weakly sclerotized. Propodosoma, Dorsum. Peritremes guttate, separated by much less than one of their Jengths, only slightly diverging anteriorly; esophageal “finned” structures indistinct. Venter. Anterior ventral plate with five pairs of setae, ventrites II] with two pairs; external ventrals I not forked. Hysterosoma, Dorsum. First segment meeting posterior margin of propodosoma, fourth tergum with but a single pair of setae. Venter. Posterior ventral plate bipartite, divided transversely be- tween hind margins of third coxae, each plate thus bearing one pair of legs; posterior ventral plate with five pairs of setae, one pair of poststernals lacking; posterior margin of posterior ventral plate with median lobe; apodemes III distinct only laterally; apodemes IV arcuate, weak but complete, arising laterally from anterior sternocoxal condyles of coxae IV; apodemes VY absent; posterior median apodeme present only on hind portion of posterior ventral plate, its anterior extremity meeting apodemes IV; extent 106 THe UNIVERSITY SCIENCE BULLETIN of pleated conjunctival bands not ascertainable because specimens are gravid. Legs. Short; leg I much thicker than leg I; tarsus I quadrate- tapering; trochanter I with four setae; femurogenu II with one seta, femurogenu III with one; tarsi IT and III each with six tactiles; solenidium of tarsus II near basal margin; pulvilli Ii to IV flaplike, longer than claws; claws II to IV stout; coxa II elongate-triangular in inner ventral aspect; coxa IV _ triangular, shorter than coxa III; femurogenu IV with one seta; tarsus IV re- sembling tarsus III in size and shape; ratio of femurogenu IV to tarsus IV approximately 1:2.7. DESCRIPTION OF GRAvID FEMALE. As described above except that body becomes physogastric. Although the type females of S. macer are gravid, it is difficult to ascertain whether physogastry involves the entire hysterosoma or whether it is of the type found in Pyemotes in which only the opisthosoma becomes enlarged. Hasits. Since gravid females were taken from the stems of sugarcane, it is possible that these mites are fungivorous in the manner of Siteroptes sensu stricto. Sporelike artifacts were found clinging to the bodies of both specimens of S. macer. The speci- mens of aradii, found upon the bodies of Aradus depressus, are the only forms known in the genus to practice phoresy. Species thought to belong in Metasiteroptes: Paradii (Krezal, 1959) ( Pyemotes ) macer, n. Sp. Siteroptes (Metasiteroptes) macer, new species Diacnosis. Separable from aradii (Krezal) in being more slender, in the size and location of the dorsal setae of all segments, in the setation (and character?) of the posterior ventral plate, and in various other characters. DescriPTION OF GrRAvID FEMALE. (Only gravid females are known.) Distance between anterior sternocoxal condyles of coxae III, 43p. (44). Gnathosoma. Elliptical, length, 20u.; width, 16y.; external dorsals absent; solenidia I indistinct, solenidia II elongate, weakly clavate; first palpal setae well anterior. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets 25y. (24); peritremes guttate, only slightly divergent, 0.4u. (0.4) apart; propodosomal setae stout, aciculate, indistinctly THe Famity PYEMOTIDAE 107 spiculate; stigmatal setae about one and one-quarter times as long as, anterior pseudostigmatals subequal to, and posterior pseudostig- matals about two and one-quarter times as long as pseudostigmata. Venter. All setae aciculate, nude, short, all those of ventrites I and internal ventrals II approximately 7-10y.; extenral ventrals II about 15y. long, distinctly behind internal sternals II and closer to coxa II than to them. Hysterosoma. Dorsum. All dorsal setae aciculate, indistinctly spiculate; dorsals I stout, 40y. long (44), well posterior to laterals I, latter about 42» (37) long, arising from posterior one-fifth of seg- ment; dorsals II as stout and long as laterals I, arising directly posterior to dorsals I; dorsals III barely median of dorsals II and as stout and long, laterals III above and laterad of these, arising from middle of the segment, approximately 14y. long; dorsals IV directly behind dorsals I and II, about as stout and long as former, laterals [IV absent; segment V with only two pairs of apicoventral setae, external pair as stout as setae of IV but shorter, about 29y. (31), internal pair about three-fifths of distance between external pair and longitudinal median line, short, 4p. (6). Venter. Posterior ventral plate elongate, distance from its anterior margin to anterior condyles of coxae HII greater than length of coxae III; posterior margin of coxae III separated by approximately 10y. from anterior margin of coxae IV; posterior ventral plate divided into a posterior and an anterior portion separated by conjunctival area, each portion bearing a pair of legs; hind margin of anterior portion rounded, meeting anterior inner margins of coxae III laterally, this portion bearing internal and external presternals and first axillary setae as well as apodemes III; hind margin of posterior portion of plate with median lobe, this portion bearing one pair of poststernals, the second axillary setae, and apodemes IV; apodemes III developed only laterally, longitudinal median suture lacking between them, apode- mes IV joining medially, forming a broadly inverted U, their point of fusion marking anterior extent of median longitudinal suture which extends posteriorly to posterior margins of coxae IV; first and second axillaries arising from their respective apodemes; one pair of poststernals, these approximately 3». behind posterior margins of coxae IV, about as close together as internal presternals; all ventral setae short, smooth and aciculate, either about 0.25 or 0.70 times length of pseudostigmata. 108 THe UNIVERSITY SCIENCE BULLETIN Legs (all measurements in»). Length. Leg I, 56; Leg II, 51; leg III, 55 (59); leg IV, 62 (63). Width. Leg I, 12 (13); leg II, 10 (11); leg III, 10 (10); leg IV, 9 (11). Segment lengths. tr I, 22 (24): ti I, 13; ta I, 12: ta I, 17: ti PM, 9 (11) cxalVee2 4 ate 24 (23); fege IV, 8 (7); ta IV, 19 (20). Setation. Leg I: ex IT, tr 4T, fege 3T, ti 6T + 2S, ta 6T + 7S; leg II: cx IT, tr 3T, fege IT, ti 4T + 1S, ta 6T + 1S; leg III: ex IT, tr 2T, fege IT, ti 4T 4+ 1S, ta 6T; leg IV: cx IT, tr 2T, fege IT, ti 4T, ta 6T. Tarsus I distinctly thinner than tibia I; solenidium 1 of tibia I stoutly pedicellate, 6.5y long; solenidium of tarsus II small (0.3y.), clavate; coxa III long-triangular, length 24, (26), width ly (1), coxa IV also triangular, shorter, length 21y, width lp (1). DisTRIBUTION. Known only from the type locality. Type MarertaL. Holotype and one paratype, both gravid fe- males, Barbados, British West Indies (intercepted in quarantine at New York, April 2, 1939, Post, collector), on stem of sugarcane, U.S. National Museum Lot Number 39-5651. Type Reposirories. Holotype and one paratype in the U. S. Na- tional Museum. Subgenus Siteroptes, sensu stricto (Figs. 10-12, 25-28). Siteroptes Amerling, 1861, Lotos, XI: 26, sine typo. Type, Siteroptes cerealium Kirchner, 1864, Lotos, XIV:126, by subsequent designation, Kirchner, 1864. Thermismoptes Amerling, 1861, Lotos, XI: 26, sine typo. Type, T. cerealium, Kirchner, 1864, Lotos, XIV: 125, by subsequent designation, Kirchner, 1864. Pediculopsis Reuter, 1907, Festschrift fiir Palmen, 7: 3, Helsingfors. Type, Pediculoides graminum Reuter, 1900, by original designation. Type: Siteroptes cerealium Kirchner, 1864, by subsequent designation, Kirch- ner, 1864. This genus contained only the species cerealiwm Kirchner, 1864 (= P. graminum Reuter, 1900) until 1957, at which time Krantz described S. reniformis. The author has at least five additional undescribed species. Members of this subgenus are considered to be more specialized than those of Metasiteroptes in that coxa IV differs in shape from coxa III and the posterior ventral plate is entire. Many members of the group have lost at least one pair of setae on the posterior ventral plate, a loss often correlated with the loss of the third laterals. Apodemes III are usually distinct only laterally and are incomplete, apodemes IV are weak, incomplete or complete but always present near the midline. Apodemes V are only rarely present. Ventrites II always bear two pairs of setae. THE FaMILy PYEMOTIDAE 109 The peritremes are only rarely guttate and if so, then are widely spaced; they usually diverge less than 90 degrees. Members of this subgenus are considered to be less specialized than those of the following subgenus although intergrades between the two groups occur. DEsCRIPTION OF NON-GRAVID FEMALE. Agrees with the descrip- tion of the genus except for the following: length, 160 to 265y, width, 75 to 130x. Propodosoma. Dorsum. Peritremes usually large, elongate-gut- tate or elongate-cylindrical, or if rarely circular or short-guttate then separated by at least one (usually more) of their own lengths, their axes usually not diverging more than 90 degrees. Venter. Anterior ventral plate with five pairs of setae; ventrites I] always with two pairs. Hysterosoma. Third tergum often with only a single pair of setae, the laterals absent. Venter. Posterior ventral plate entire; apodemes III incomplete medially, usually not extending mesally beyond internal presternals; apodemes IV weak, complete or in- complete, always distinct along midline and laterally in front of coxae IV; apodemes V usually absent; posterior median apodeme weak or absent anterior to its union with apodemes IV. Legs. Trochanter I with four setae; femurogenu II with one to three (usually two) setae; tarsi II and III each with seven tactiles. Species thought to belong in Siteroptes s. str.: antiquissimus (Krezal, 1959) (Pygmephorus ) avenae (Miiller, 1905) ( Pediculoides ) cerealium Kirchner, 1864 dianthophilus (Wolcott, 1908) (Pediculoides ) graminum (Reuter, 1900) ( Pediculoides ) primitivus (Krezal, 1959) (Pygmephorus) reniformis Krantz, 1957 Hasits. Siteroptes cerealium * is important as a principal vector of Fusarium poae (Peck) (Deuteromycetes), the pathogen of a disease variously known on a number of meadow grasses as “silver- top,” Weissahrigkeit, or Bollnaserkrankheit, and on carnations as “central bud rot” (Cooper, 1940). In addition, it has recently (Alfaro, 1946) been implicated in the transmission of Nigrospora oryzae (Berk. & Br.), a fungal parasite of grains. 4. A discrepancy in the number of chromosomes found by Reuter (1909b) and Cooper (1987) was found to exist. Cooper (1939) states, after examination of some of peters slides, that the latter author was in error, the correct diploid chromosome number eing six. 110 THE UNIVERSITY SCIENCE BULLETIN The most important works dealing with the bionomics, genetics, morphology, and embryology of this species are those of Alfaro (1946), Bradbury (1926), Cooper (1937, 1939, 1940), Molz and Morgenthaler (1912), Reuter (1900, 1907, 1909), Stewart and Hodgkiss (1908), Weiss (1915), and Wolcott (1908). The following account of the life history of S. cerealium is taken largely from Reuter (1900 and 1909a), and Cooper (1937, 1940). Apparently, only the young mated females of this species dis- perse. These females readily leave their place of birth and seek food and shelter. They are capable of traversing tight places and are able to escape from and enter vials tightly stoppered with cotton plugs. Those attacking grasses commonly take up residence be- tween an upper (usually the uppermost) leaf sheath and the culm; those attacking carnations force their way into the buds where they are found near the inner floral organs. Once so located, the young female begins to feed, this feeding being prerequisite to physogastry. The exact nature of the food is not known. All authors agree that tissue necrotized through the invasion of F. poae is ingested. Alfaro (1946) and Reuter (1900) state, in addition, that the mites are capable of attaining physogastry after feeding upon healthy tissue. Cooper (1940) doubts this to be the case, finding physo- gastric females to be associated only with diseased plants in every instance. The same author found that the mites will feed and readily subsist upon the mycelium of F. poae isolated in agar plates. The duration of the feeding period before the onset of physo- gastry is not clear but Cooper (1940) mentions one case in which noticeable swelling occurred six days after the females were placed on the soil near carnations in a greenhouse. How much, if any, of this period was spent in waiting for the plant tissue to become suitably decomposed by the fungus is unknown. Females become increasingly less mobile as physogastry pro- gresses and finally become totally incapable of movement, remain- ing anchored by palps and (probably) by at least one pair of front legs until they die. The volume of these females may be increased to 100 to 500 times that of non-gravid forms, and their width may measure 3000u.. Like those of many other genera and unlike those of Pyemotes, the entire hysterosoma swells, balloonlike, while the propodosoma remains unmodified. Within this fluid-filled hystero- somal globe, the eggs develop and hatch into active hexapod larvae. In most instances, these remain within the maternal body, where THe Famity PYEMOTIDAE ela they supposedly quickly molt to the octapod nymph. According to Reuter (1900), these latter are distinguishable as to sex and differ from the adult only in the development of the organs of reproduction. The appearance of this stage is an anomaly seem- ingly restricted to members of this genus, or perhaps to cerealium. I have seen males and larvae of several species but no nymphs. Perhaps later research will prove Reuters nymphs to have been merely teneral adults. The species differs from previously discussed members of the Pyemotinae and from known Acarophenacinae and resembles most Pygmephorinae in having at least one definitive immature stage. According to Reuter, larvae and nymphs (especially the latter ) are often born and move freely about the host plants where they soon settle and molt to the ensuing stages. Cooper (1937) states that this is rarely the case and that usually only adult mites are born. The above author gives the time of development from egg to nymph as four to five weeks but makes no mention of the tempera- tures at which this rate of development occurred. Both he and Alfaro state that two or three generations occur in a season. Although Reuter thought that mating takes place on the host plant, Cooper (1937) found that laboratory-reared mites usually mate in the adult stage while still within the body of the mother but noted that in rare instances mating took place while adults were still upon the surface of the maternal hysterosoma. Copula- tion was observed to last between 5 and 35 seconds. The mass birth of the offspring is attended by the hysterosomal breakdown and accompanying death of the mother. Cooper (1937) counted the offspring obtained from three series of females. The first series contained 17 female parents and totalled 1643 offspring (mean, 96.5/female), the second 72 females and 11,446 offspring (mean, 158.8 offspring/female), and the third 30 females and 4385 offspring (mean, 146.0/female). No information as to the circumstances surrounding the procurement of these three samples or reasons for the large differences in mean fecundity be- tween them are given. The same author found males to occur in every brood, although in small numbers. The three samples men- tioned above produced 4.50, 4.46, and 5.15 percent males respec- tively. Reuter (1900) found no males in the spring but a progres- sive increase in their numbers during the summer until a high of about 13 percent males was reached in the fall. This increase was thought to coincide with the need for relatively large numbers of 112 THe UNIVERSITY SCIENCE BULLETIN mated overwintering female nymphs and the absence of males in the spring was attributed to the fact that they do not overwinter. On the basis of the findings made in related genera, it is most probable that a relatively constant number of males is produced in every generation, as inferred by Cooper. Similar evidence sug- gests that mated females, rather than nymphs, are the overwinter- ing forms. The above two authors concur in recording high mortality in- variably occurring in unmated females. Cooper noted only two cases in which such females became physogastric. Both were arrheno- tokous. It is interesting to note that the mean number of offspring per female (306) in these two cases was considerably higher than in mated females. Perhaps this discrepancy may be explained by the relatively smaller size of the male. Gravid females may be found in gradually increasing numbers until October. These produce large numbers of forms which (see above) Reuter believed to be overwintering female nymphs. Ac- cording to that author, such nymphs may be virgin or may be mated immediately after birth. The former usually die the ensuing spring without molting to adulthood. Cooper (1940) has demonstrated the ability of S. cerealium to transmit the spores of F. poae to carnation, and Alfaro (1946) has experimentally transmitted the fungus Nigrospora oryzae by means of the same mite. The present author has sometimes seen elliptical or circular sporelike objects upon or within the bodies of Siteroptes females. The relationship between the fungus and the mite is stated to be a symbiosis in which the mite disseminates and innoculates the fungus which alters the plant tissue to provide (more suitable) food for the mite. Cooper points out that this symbiosis is antago- nistic in that the mite may feed upon the fungus. The phenomenon investigated and named Karyomerokinesis by Reuter (1909b ) is reinvestigated in detail by Cooper (1939). Aside from a unique elaboration of Feulgen-negative chromosomelike material between separating chromosomes, mitosis in the early cleavage divisions of the egg is similar to that found by Pitau (1936) for Pyemotes ventricosus. Karyomery and the formation of equatorial bodies ceases by the tenth cleavage division. Karyomeres disappear in the prophase, reappearing as vesicles in the telophase. Specimens examined. Nine species, including reniformis and prob- THE FamMity PYEMOTIDAE 113 ably cerealium and primitivus, from many parts of the United States and from France, from the following situations or hosts: Salt marsh grass (Distichlis sp.), grass, Celtis occidentalis L., carnations, cotton bolls (syntypes of reniformis), milo maize, Setaria sp., nest of Microtus sp., treehole in Fagus grandifolia Ehrh., peach orchard soil, decaying log, forest floor litter. Siteroptoides, new subgenus (Figs. 51-53 ) Type, Siteroptes (Siteroptoides) absidatus, new species. This new subgenus contains at least six species. Although its forms intergrade with those of Siteroptes s. str., it is considered to be more specialized than the latter and perhaps more closely related to Pediculaster and the Pygmephorinae, as evidenced by the develop- ment of the apodemes of the posterior ventral plate, by the loca- tion, proximity, and amount of divergence of the peritremes, and by the additional pair of setae of the second ventrites. DescrIPTION OF NON-GRAVID FEMALE. Length, 150 to 320y. width, 80 to 160y.; hysterosoma elliptical. Propodosoma. Dorsum. Peritremes guttate or elongate-guttate, occasionally septate, usually diverging at least ninety degrees, separated at their bases by less than one of their widths. Venter. Anterior ventral plate with five (rarely) or six pairs of setae, when six then ventrites II with three pairs. Hysterosoma. Dorsum. Third and fourth terga usually with two pairs of setae, laterals rarely reduced or absent. Venter. Posterior ventral plate entire, usually bearing six pairs of setae, or (rarely ) internal poststernals absent; external poststernals always well be- hind posterior margin of coxae IV; apodemes III always well developed, complete or nearly so, bending posteriorly at the sides to join anterior sternocoxal articulation of II, bending posteriorly in the middle to join posterior median apodeme; apodemes IV complete (sometimes distinctly weakened near posterior median apodeme ), usually indistinctly forked laterally, one branch bending posteriorly to join anterior sternocoxal articulation IV, the second branch meeting and continuing posteriorly and dorsally around the margin of coxal foramen III; apodemes V well developed, joining the posterior extremity of the posterior median apodeme; posterior median apodeme always well developed anterior to apodemes IV, meeting apodemes III. 114 THe UNIVERSITY SCIENCE BULLETIN Legs. Trochanter I with four setae; femurogenu II always with three tactiles; tarsi If and III each with seven tactiles. Species thought to belong in Siteroptes ( Siteroptoides ): absidatus, n. sp. priscus (Krezal, 1959) (Pygmephorus ) Hasits. Nothing is known of the habits of this group. Those in the author’s collection were collected in a variety of habitats—on plants, in soil, in a treehole, and in cattle manure in a rearing cage of Siphona irritans. One specimen (from Soledad, Cuba) has a single sporelike structure attached to it. SPECIMENS EXAMINED. In addition to the type series of absidatus probably six species from many parts of the United States and from Cuba, Haiti, and Holland, from the following situations or hosts: treehole in Fagus grandifolia, stems of Paeonia sp., corm of Gladi- olus sp., oats, and on cattle droppings in cage of Siphona irritans L. (Diptera). The specific status of some of the specimens is not clear. Includ- ing absidatus, the author has three forms which are closely related and which may be conspecific, differing in size, shape, and location of the solenidia of tibia and tarsus I as well as in minor features of the placement and size of the setae of terga HI, IV, and V. In view of the small number of specimens at hand, obviating any possibility of ascertaining intergradation, if present, it has been con- sidered best to restrict the definition of absidatus to the three speci- mens described. Siteroptes ( Siteroptoides) absidatus, new species Diacnosis. Differs from priscus (Krezal) in the shape of soleni- dia 3 and 4 of tarsus I, in that the peritremes are more widely spaced, in that external caudals II are much longer than the other two pairs, and in various other characteristics. Description OF Non-Gravip FemMaAte. Length, 265y, (150-265); width, 132y, (80-132); distance between anterior sternocoxal con- dyles of coxae III, 74. (55-74). Gnathosoma. Quadrate, approximately 30 x 30u; external dorsals arising well laterad of and slightly behind internal dorsals, aciculate; solenidia I distinct, appearing concave apically, solenidia II elon- gate, clavate; distance between first and second palpal setae about 7y; esophagus with small, slightly thickened mesal dilation. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets 29y, (22-30); peritremes guttate, greatly divergent, THe Famity PYEMOTIDAE 115 their long axes making an angle of about 130 degrees, their most mesal margins 4y, (2-4) apart; propodosomal setae stout, aciculate, spiculate, the stigmatals 29». (17-29), the anterior pseudostigmatals only slightly shorter, the posterior pseudostigmatals comparatively short, 40y. (29-42), arising slightly behind internal pseudostigmatal sockets and on longitudinal line drawn through their mesal margins. Venter. Six pairs setae, ventrites I and II each with three pairs; setae of anterior ventral plate aciculate, nude, short, range 11 to 20u, median ventrals I closer to anterior median apodeme than median ventrals II, external ventrals I not forked, arising posterior to internal ventrals I; ventrals II subequally spaced and sublinearly arranged. Hysterosoma. Dorsum. All dorsal setae stout, aciculate, sparsely spiculate; setae of segment I in a transverse row, dorsals I 29u. (18-30), 40u. (30-42) apart, laterals I 41y. (26-45), distinctly closer to respective dorsals than dorsals to each other; dorsals II 34u. (24-39) long, 50. (39-53) apart; dorsals III as laterals I, 46y, (33- 52) apart; laterals III 19». (9-19) arising posterolaterad (not more than 20, laterad) of dorsals; dorsals IV as laterals I, 32y, (20-33) apart, laterals IV 19. (15-22) long, distinctly farther from their dorsals than those of third tergum; external caudals II 7-22y. long, distinctly longer than other two pairs; apical portion of vagina with distinct sclerotized thickening that is globular to elliptical in shape. Venter. Hind margin of posterior ventral plate broadly and shallowly emarginate; apodemes III mesad of internal pre- sternals and posterior median apodeme between apodemes III and [V weakly developed; internal presternals arising from apodemes III, external presternals arising about two areolar diameters an- terior to apodemes IV, directly behind internal presternals; internal poststernals present, distinctly mesad of external presternals, ex- ternal poststernals arising behind posterior terminations of apodemes V, laterad of external presternals; second axillaries immediately mesad of apodemes IV, slightly anterior to internal poststernals; all setae of posterior ventral plate flagellate, nude, similar in shape, short (7-18y.), longest on any one specimen never twice length of the shortest. Legs (all measurements in »). Length. Leg I, 90 (65-94); leg IT, 72 (52-77); leg III, 72 (51-80); leg IV, 105 (69-110). Width. Leg I, 17 (13-18); leg II, 13 (9-13); leg III, 11 (8-12); leg IV, 11 (8-12). Segment lengths. tr I, 33 (24-36); ti I, 20 (14-22); ta I, 95 (16-25); ta II, 24 (16-25); ti III, 13 (8-14); cx IV, 30 (19-34); 116 THe UNIVERSITY SCIENCE BULLETIN tr IV, 26 (17-28); fege IV, 8 (6-9); ta IV, 32 (19-33). Setation. Leg I: cx IT, tr 4T, fege 4T, ti 6T 4+ 2S, ta 15 (T and S not sepa- rable in most instances): leg II: ex 1T, tr 3T, fege 3T, ti 4T + 1S, ta 7T + 1S; leg III: cx IT, tr 2T, fege 2T, ti 4T + 1S, ta 7T; leg IV: ex IT, tr 2T, fege IT, ti 4T +15, ta 6T. Tarsus I with only two clavate or cylindrical solenidia, the rest setalike; solenidium 3 rodlike, curved, slightly longer than claw, arising at level of pedicelar base; solenidium 4 rodlike, straight, less than half as long or wide as 3, arising behind 3 but in apical third of segment; solenidium of tarsus II thickly rodlike, elongate, basal; solenidia of tibiae II and III small, clavate, arising at basal quarter of segment; tarsi II to IV constricted suddenly near most apical setae; coxa IV wider at base than at apex; solenidium of tibia IV small, weakly clavate, arising at basal quarter of segment; pulvilli II to IV shorter than respective claws; ratio of femurogenu IV to tarsus IV, approximately 1:3.1 to 1:4.0. VaRIATION. Little variation in setal placement, size, etc. is found in the type series. A great range in size exists, however, as is in- dicated by the presence of one very small specimen. The measure- ments of this specimen are available from the description since they are the smallest measurements in every case. DiIsTRIBUTION. Known only from the type locality. Type MaterraL. Female holotype and two female paratypes (none gravid) from Howard County, Arkansas, June 8, 1936, W. F. Turner, collector. U.S.N.M. Lot Number 36-20984. Typrr Repositories. Holotype and paratype No. 2 to the United States National Museum. Paratype No. 1 to the Snow Entomologi- cal Museum, The University of Kansas. Trochometridium, new genus (Figs. 29-40) Type, Trochometridium tribulatum, new species. This genus is erected to accommodate a single new species, which while showing affinities to both Siteroptes and Pyemotes, is very distinct from either. It resembles the former and differs frem the latter in having a squarish gnathosoma bearing free palpi and dor- sal, approximate stigmata, in that coxae IV are quadrate (although distinctly dicondylic), in the structural conformation of the gravid female, in that mass laying of eggs (accompanied by hysterosomal breakdown) occurs, and because the male mouthparts are vestigial. THE FAMILY PYEMOTIDAE 117 It resembles Pyemotes and differs from Siteroptes in food habits, in possessing stout, easily distinguished chelicerae, in possessing a tripartite posterior ventral plate, in lacking a distinct posterior median apodeme, in that claw I of the female is sessile, and because leg IV of the male is furnished with two claws. It resembles neither of the above in that the first hysterosomal segment is tri- partite, in its mating habits, in the sex ratio, in that the female trochanter I bears five setae rather than four, in the genital mor- phology of the male, and because leg I of the male is clawless. In consideration of the above, the genus seems clearly to belong in the Pyemotinae as a specialized form closely related to the Acaro- phenacinae. Dracnosis. The females are easily separable from those of all other genera in that legs I are five-segmented, the recurved chelic- erae are stout and distinct, the palps are free, and because the first hysterosomal segment is divided into one median and two lateral plates. Additional distinguishing characters are the presence of paired germaria between legs II and III, the dorsal, approximate, capitate peritremes, and the tripartite posterior ventral plate. Males are distinguished from those of other genera in that the head, while reduced and lacking visible mouthparts, is bilobate rather than peglike, because leg I lacks claws but possesses a flap- like apical dorsal pulvillus, in that leg IV is more slender than legs I and IJ and bears two unequal claws (the smaller half the width of the larger), in that the aedeagus is ventral, short, bipartite, at- tenuate, declinate, non-cylindrical, and uncoiled, and because the opisthosoma is not modified (conelike or globelike ) to accommodate the genital and clasping structures. DerscriPTION OF NON-GRAVID FEMALE. Length, approximately 2504; width, approximately 150u; width between anterior condyles of coxae III, approximately 115y; hysterosoma elliptical-fusiform to elliptical, its posterior margin evenly rounded; body moderately sclerotized. Gnathosoma. Squarish-fusiform, normally directed anteriorly, length approximately equal to width; palp free, distinct, with single small apical tooth; chelicera stout, styliform, recurved, easily visible; esophageal-pharynegeal pump absent. Propodosoma. Dorsum. Inversely T-shaped, posterolateral mar- gin rounded, anterior margin forming short, cylindrical cervix, pos- terior margin overlapping first hysterosomal segment; pseudostig- 118 THE UNIVERSITY SCIENCE BULLETIN mata located in anterior one-half; three pairs of propodosomal setae; peritremes dorsal and median, close together, nearly trans- verse, capitate medially; stigmatal openings arcuate, slit-like; esoph- ageal structures not visible. Venter. Anterior ventral plate with six pairs of setae, ventrites I and II each with three pairs; external ventral I not forked, medial ventrals Il much closer to externals II than to internals II; coxa I attached at nearly a right angle to longi- tudinal axis of body; apodemes II very strong, straight, complete medially, making angle of approximately 45 degrees with anterior median apodeme; anterior median apodeme extremely large, com- plete; posterior marginal apodemes weakly developed, distinct only laterally; pleated conjunctival band present between propodosoma and hysterosoma. Hysterosoma. Dorsum. First dorsal segment divided into one median and two lateral plates, each lateral plate bearing one seta, the median plate bearing a pair; fourth segment with two pairs of setae; fifth segment well developed but usually completely en- closed by fourth in dorsal aspect, bearing two pairs of setae; vulva apicoventral, its margin apparently greatly thickened and sclero- tized; one omega-shaped internal genital structure always distinct (fig. 30). Venter. Posterior ventral plate tripartite, the two lateral portions overlapping each other mesad to coxae LI, the median portion visible posteriorly and mesally as a triangular area beneath and between margins of lateral plates; posterior ventral plate with six pairs of setae; first axillary setae well anterior and mesad of anterior sternocoxal condyles of coxae III; second axillaries well inside and slightly anterior of coxae IV; external presternals well behind and well laterad of internal presternals, arising nearly half way between internals and first axillaries; external poststernals pres- ent, on median portion of posterior ventral plate; apodemes III not developed except at anterior sternocoxal condyles III, apodemes IV weak, joining coxae IV at anteromedial corners, progressing mesally and anteriorly to join with apodemes V, the latter joining coxae [V at posterolateral corners and proceeding anteromedially; posterior median apodeme absent; developing eggs in linear series sometimes visible in germaria situated laterally between legs II] and IV; band of opisthosomal conjunctival pleats not visible on all specimens; opisthosomal venter entire, bearing a single pair of setae. Legs. Well developed and long, margins of segments often with sclerotized areas of varying thicknesses; leg I five-segmented, about THe Famity PYEMOTIDAE 119 twice as wide as leg II; tarsus I subconical, not wider than femuro- genu I; large, sclerotized areas present along inner margin of tarsus — I and medially near outer margin of tibia I; claw I single, stout, sessile, furnished with two stout thumbs; trochanter I with five setae, seta c long, aciculate, sparsely plumose, reaching beyond tip of tarsus; seta d hooked; sensory pegs of tibiae II and III present; tarsi II and III with eight tactiles; claws Il to IV simple, subequal, a single short spur arising in apposition to them at apex of tarsus; pulvilli IL to IV distinctly longer than claws; coxa HII triangular; coxa IV quadrate, distinctly mesad of III; tarsus IV moderately long, tapering evenly, ratio of femurogenu IV to tarsus IV approxi- mately 1:2.5. DESCRIPTION OF GRAVID FEMALE. Gnathosoma and propodosoma as described for non-gravid female but entire hysterosoma becoming physogastric, globular, 2 to 3 mm. in diameter. DrEscrIPTION OF Mater. Length, approximately 250»; width, ap- proximately 170u; hysterosoma broadly oval, not sharply constricted in vicinity of legs III; conjunctiva granulose, especially laterally and ventrally between anterior and posterior ventral plates. Gnathosoma. Length, approximately 20%; width, approximately 35u.; reduced, bilobate, palps lacking, oral appendages and pharyn- geal pump not visible, probably lacking as well. Propodosoma. Dorsum. Dorsal plate well defined, short, sub- quadrate, enclosing reduced gnathosoma dorsally and laterally, its posterolateral margin bearing three pairs of setae. Venter. Apo- demes I, I, and anterior median apodeme greatly enlarged, well sclerotized; anterior ventral plate with six pairs of setae, ventrites I and II each with three pairs. Hysterosoma. Dorsum. Three dorsal plates present, the first bearing three pairs of setae, the second two pairs, and the most caudal two pairs. Venter. Posterior ventral plate entire; apodemes III, IV, and V well developed, joining equally distinct posterior median apodeme; external presternals well behind and outside of internal presternals; external poststernals present, slightly inside of and well behind internal poststernals; hind margin of posterior ventral plate indistinct; posterior margin of hysterosomal venter indistinct or distinct; posterior margin of opisthosoma sharply rounded; the halves of the apical aedeagus joining to form a short armature, the latter hooked ventrally near its apex and usually pro- truding slightly from the body; well defined genital capsule or spe- 120 THE UNIVERSITY SCIENCE BULLETIN cialized clasping structures lacking; two pairs of spinelike apical setae flanking aedeagal fissure, one pair of longer setae well ventrad and laterad of these. Legs. All legs well developed, five-segmented; tarsus I lacking claw, with flaplike dorsal pulvillus; leg IV not greatly modified in size or form; coxa IV subquadrate; trochanter IV not distinctly arcuate; tarsus IV well developed, with two claws, one similar in size to claws of other legs, other twice size of first. Trochometridium tribulatum, new species DESCRIPTION OF NON-GRAVID FEMALE. Body elliptical-fusiform, length, 247y, (218-295, X = 253.08 + S. E. 6.59); width, 146y, 127- 187, X = 153.63 + S.E. 4.88); distance between coxae III, 115v. (108-124, X —115.22 + S.E. 1.78); at least posterior portion of propodosomal dorsum and median portion of first dorsal segment coarsely punctate. Gnathosoma. Squarish-fusiform, large, approximately 48 x 48u; all gnathosomal dorsals present, internals tiny, arising anterior to and slightly laterad of externals; solenidium II tiny, appearing as a palpal tooth; solenidium I subapical, lanceolate; one tiny seta present on oblique inner face of palpal apex; second palpal seta one-third of length of palp from apex, distance between first and second palpal setae about 7y; esophageal pump not visible. Propodosoma. Dorsum. Width between pseudostigmatal sockets 69u., (63-70, X = 67.00 + S. E. 0.88); peritremes nearly transverse, irregularly shaped, capitate medially, about 20y. center-to-center, most mesal margins about 3v. apart; stigmatal openings slitlike, arcuate; pseudostigmata barely in anterior half of propodosoma, line drawn between them about 5y. in front of stigmatal setae; propodosomal setae variable, sparsely short-plumose, stigmatals di- rectly below median third of peritremes, about 26y long; anterior pseudostigmatals short, about 14y, arising mesad and posterior to internal pseudostigmatal sockets; posterior pseudostigmatals about 70u. long. Venter. Six pairs setate, ventrites I and II each with three pairs; median and internal ventrals II longer than rest, 40 to 50u, internal ventrals I shortest, 12 to 17; internal ventral II well anterior to other setae of that ventrite; hind margin of anterior ventral plate irregularly scalloped, not heavily sclerotized, posterior marginal apodemes absent except at sides. Hysterosoma. Dorsum. Dorsal setae varying in size, always sparsely plumose; anterior portions of all three sections of the first THe Famity PYEMOTIDAE 121 segment covered by posterior margin of propodosoma; dorsal setae I well posterior to laterals I, arising from posterior third of segment, stout, 41u, 38-43, X = 39.43 + S.E. 0.40); laterals I longer, about 75u, otherwise similar to dorsals I; dorsals II as stout as dorsals I and longer, 48y., (41-48, X = 44.57 -- S. E. 0.68) slightly laterad of latter, spaced about 36y. apart; all setae of segment III equally short, 17p (17-20, X = 18.21 + S. E. 0.35), the dorsals about 9y. apart and slightly anterior to laterals; dorsals IV approximately 20y. apart, 5ly. (43-54, X = 48.14+ S.E. 0.99) long, stout, formed as setae of segment I, laterals IV on transverse line with dorsals IV, only a third their length; segment V with two pairs of short, weakly plumose ventral setae about 12 in length, first pair flanking the vulva, second pair nearly a setal length anterolateral. Venter. Lateral divisions of posterior ventral plate arcuate along their mesal margins, these margins overlapping each other in the vicinity of legs III; lateral portions of posterior ventral plate bearing all setae except external poststernals, these borne on triangular median portion of posterior ventral plate (fig. 30); anterior condyles of coxa III well sclerotized, but apodemes III absent; apodeme IV joining apodeme V mesally, delimiting a bell-shaped area which bears the second axillaries and internal poststernals; posterior median apodeme indistinct (distinct as short apophysislike invagina- tion); first axillary well anterior and mesad of anterior condyle of coxa IIT, second axillary immediately anterior and mesad of anter- omedian corner of coxa IV; internal poststernals well (about 10y.) anterior to anterolateral corner of coxa [V, about one and one-half times as far apart as internal presternals; all setae of posterior ventral plate thin, flagellate, sparsely spiculate, 25 to 35y. long; opisthosomal venter distinct, cleft medially, cleft often appearing to contact internal genital sclerotization; setae of opisthosomal venter approximately 15y, long, about as far apart as internal ventrals II. Legs (all measurements in y.). Length. Leg I, — (98-122, K = 113.50=+S. E. 2.99); leg If, 113 (110-118, KX = 113.90 =S. E. 0.78); leg III, 125 (125-128, X = 125.86 +S. E. 0.44); leg IV, — (139-152, X= 14517458. EF. 1.82). Width. leg I, 31 (22-32, K = 28.92 + S. E. 0.79); leg II, 19 (14-21, KX = 17.84=S. E. 0.47); leg III, 18 (13-18, X = 16.61 +S. E. 0.50); leg IV, 20 (13-20, KX = 18.08 +S. E. 0.58). Segment lengths. tr. I, 42 (40-43, X = 41.78 +S. E. 0.44); fege I, 22 (20-238, KX = 2144+S. E. 0.58); ta II, 35 5. Measurement not possible on holotype. 122; THe UNIVERSITY SCIENCE BULLETIN -39) X= 36/67 = S: FE. 0:65) fege I, 19° (18-21 eager 82); tr IV, 41° (41-45, X = 42.66 = S. E. 0.42); ex IV, 23 (22- D, X == 23164 = S. E: 0:25); fege 1V, Qik (19-22) xe 20a iorele ); ta IV, 52 (48-56) X = 53.33 + S.E. 0.72). Setationn Meow: cx IT, tr 5T, fege 5T, ti 6T + 2S, ta 71 + 6S; leg II: cx 1T, t 30, fege 3T, ti 4T + 1S, ta 8T + 1S; leg III: cx IT, tr 27, fege 37, ti AT + 1S, ta 8T; leg IV: cx 1T, tr 2T, fege 2T, ti 4T + 1S, ta 7T. Trochanter I with five setae, the most basal and medial hooked, the apical external dorsal long and whiplike, sparsely plumose, reaching apex of leg or beyond; tibia I distinctly wider than any other leg segment; tarsus I tapering, subconical; solenidium of tarsus II bearing an apical prickle; solenidia of tibiae II and II thin, seta- like, 7 to 10u long, solenidium of tibia IV setalike, longer, 13 to 15y; coxa IV subquadrate; tarsus IV evenly attenuate, width at apex about three-fourths that of base; ratio of femurogenu IV to tarsus IV approximately 1:2.5. DescriPpTION OF Mar. Body broadly oval; length (excluding gnathosoma) 250. (240-315); width at widest point, 170y. (varia- tion not measurable); sclerites impunctate; conjunctival areas dis- tinctly granulose. Gnathosoma. Reduced; length, 28y. (variation not measurable); width, 36, (35-38); bilobate, lobes connected anteriorly by an in- distinctly defined membrane. Setation. Three pairs of short, dorsal tactiles, one pair of short, one pair of long ventral tactiles, one pair of short, rodlike, ventral sensory setae. Propodosoma. Dorsum. Weakly sclerotized except for dorsal plate; length of dorsal plate 37y. (variation not measurable); width of dorsal plate 55y. (55-69); dorsal plate with three pairs of setae, anterior pair short, 13y (13-15), second pair longer, 2ly. (20-28), posterior pair at least thrice length of median pair, 72y. (72-105). Venter. Median and internal setae of ventrites II nude and flagel- late, 100 to 140u, external ventrals II short, weakly plumose, 18v. (16-20). Hysterosoma. Dorsum. First hysterosomal plate transversely elliptical, largest of three, overlapping second; lateral margins of second plate narrowing posteriorly, overlapping third; third plate truncate posteriorly, much smaller than second. Setation. All setae of first and second dorsal plates weakly short-plumose, setae of third dorsal plate thick and nude; most anterior dorsals I 19v. (19-38) long, posterior dorsals I 82y. (81-151), laterals I 122u. (122- THE Famity PYEMOTIDAE 123 183); dorsals I] 15y (15-26), laterals Il 53y (33-81); dorsals LI short, spiniform, divergent, 5y. (5), laterals III straight, bladelike, 15y (15). Venter. All setae of posterior ventral plate short, thin, barely spiculate, 12 to 23y long; external presternals usually at least twice as far apart as internal presternals; opisthosomal ventral setae longer and stouter than setae of posterior ventral plate, 28y. (28-46) long, about one and one-half times as far apart as internal post- sternals; aedeagus apicoventral, short, approximately 13. long, bi- partite, attenuate, declinate, uncoiled, its bulbose bases divergent within the body; aedeagus usually protruding slightly from aedeagal fissure located posteriorly in fifth segment; two pairs of caudal setae arising from margin of segment V, inner pair stout, divergent, blade- like, 4p. (4-5) long, outer pair straight, spiniform, 3p (3-4) long, a single pair of short setae about 10y long (opisthosomal ventrals? ) arising well anterior and laterad of these. Legs (all measurements in p.). Length. Leg I, 130 (127-168); leg II, 154 (144-216); leg II, 151 (149-211); leg IV, 118 (118-166). Width. Leg I, 25 (25-31); leg II, 26 (26-33); leg III, 17 (17-23); feo IVa DIE (E28). setation = Lee Wee cx Iie tr ol tese aati 6T + 2S, ta 7T + 85; leg II: cx IT, tr 3T, fege 3T, ti 41 + 1S, ta ST + 1S; leg III: cx IT, tr 2T, fege 3T, ti 4T + 1S, ta 85; leg IV: cx IT, tr 2T, fege 2T, ti4T + 1S, ta7T. Femurogenu II with a single spurlike seta, tibia II with a spur- like seta stouter than that of femurogenu; tarsus II with a pair of stout spurlike setae; trochanters I and II each with a single whip- like seta; sense pegs of tibiae II, IiI, and IV thin and setalike, 7 to 12u, long; sensory peg of tarsus II bulletlike, 6 to 12y long; apical sensory setae of tarsi IT to IV small, thin, usually difficult to see. DisTRIBUTION. Oregon to Maryland and as far south as Mata- moros, Tamaulipas, Mexico. Type MatertaL. Holotype female, allotype male, six female (nos. 1, 2, 8, 9, 11, 12) and three male (nos. 14, 15, 16) paratypes from Logan, Green Canyon) Cache County, Utah, July 31, 1956, G. E. Bohart, A. Léken, E. A. Cross, from cells of Halictus farinosus Smith.© Four female paratypes, (nos. 3, 4, 5, 10) same locality and collectors, July 18, 1956, from mesosoma of Sphecodes arvensiformis Cockerell.© One female paratype (no. 7) from Lawrence, Douglas County, Kansas, July 15, 1957, A. F. Shinn, from cell of Calliopsis andreniformis Smith.‘ One male paratype (no. 17) same locality, 6. Det. P. H. Timberlake, Citrus Experiment Station, Riverside, Calif. 7. Det. A. F. Shinn, Stephen F. Austin St. Coll., Nagadoches, Texas. 124 THE UNIVERSITY SCIENCE BULLETIN collector, and host as preceding, but dated July 27, 1957. One female paratype (no. 6), Fort Meade, Anne Arundel County, Mary- land, August 3, 1956, S. R. Joseph, from legs and metasoma of “mutillid.”. One female paratype (no. 13) 5 mi. NE Adrian, Mal- heur Co., Oregon, July 31, 1958, R. Eppley, from cell of Nomia (Acunomia) melanderi Cockerell.® Type Reposirorics. Holotype, allotype, one male (no. 14) and eight female (nos. 4, 6, 7, 9, 10, 11, 12, 13) paratypes in the Snow Entomological Museum, The University of Kansas; one male (no. 17) and one female (no. 3) paratype in the United States National Museum, Washington, D. C.; one male (no. 16) and one female (no. 2) paratype in the Zoologisches Institut der Universitat Er- langen, Erlangen, Germany; one male (no. 15) and one female (no. 5) paratype in the British Museum (Natural History ), London, England; one female paratype (no. 1) in Oregon State College collection, Corvallis, Oregon; one female paratype (no. 8) in the private collection of the author. SPECIMENS EXAMINED AND Hosts. In addition to the type material the following specimens have been examined: Iowa: Sioux City, ex Calliopsis andreniformis. NEBRASKA: Lincoln, ex Calliopsis andreniformis. TAMAuLIPAS: Matamoros, ex Nomia (Epinomia) nevadensis bakeri Cockerell. Texas: Cotulla, ex N. n. bakeri. Hasirs. As far as is known, the species is a parasite of larval bees. Observations concerning its life history were made by the author while studying a nesting area of Halictus farinosus Smith, a large, soil-nesting, social halictine bee, in Green Canyon, Cache County, Utah, during late July and early August of 1956. Several cells of this bee were found to be infested with Trochometridium tribulatum, including two that contained gravid female mites and eggs or eggs and immatures. The contents of the cells were reared in the labora- tory and observations were made on their development. The eggs are large (160-200y.), spherical, translucent to milky- white. They are deposited in bunches within the cell, apparently normally, through the vulva, rather than through the rupture of the hysterosomal wall as described for Siteroptes cerealium. It resembles this species, however, as well as members of the genus Pyemotes in the complete suppression of active immature stages. The egg hatches into a quiescent form which gradually increases in size, changes shape and color, and molts into an active adult mite of either sex. 8. Det. E. A. Cross, Northwestern State College, Natchitoches, La. THe Famity PYEMOTIDAE 125 The shell splits, disclosing the quiescent embryo enclosed within an embryonic sheath. At this point only the two anterior pairs of legs protrude, and the young embryo resembles a minute ptyctime. Embryos at this stage possess two pairs of clawless hysterosomal lobes and presumably pass the larval stage while within the chorion. As the embryo grows, it casts aside the halves of the shell and takes on a more definite shape, gradually acquiring a pale reddish tinge. Both sexes at this stage are fusiform, having the two anterior pairs of legs extended forward and touching each other apically, the two posterior pairs extending behind, also touching apically. Such embryos appear to have only one enveloping membrane. The adults emerge from this membrane head first, and mites having only the first two pairs of legs free are commonly seen scraping against the substratum to peel away the posterior half of this enveloping sac. Although it was not possible to ascertain the exact duration of the immature stages, some information was obtained. An egg mass collected on July 31 from a bee cell which also contained a second gravid female mite (this mite proved to contain no mature eggs when dissected) was observed in the laboratory. On August 2, the first egg hatched, and by August 8th most of the viable eggs had hatched. On this latter date several adult males were already present. By August 10th, both males and females were present in some numbers, males comprising about 40 to 45 percent of the total population. On August 16th, approximately three-fourths of the mites had reached adulthood, at which time the total number of adults was 210. Sixty (29 percent) of these were males. Male mortality was seen to begin within two days of the time of their emergence, and what percentage survived to contribute to the above total is not known. Observations made of infested cells shortly after the completion of mite emergence indicate that the percentage of males is very low (2 percent), or that males are absent entirely. It appears, therefore, that males emerge first and die first, their complete disappearance coinciding with the time of emergence of the latest females. The male/female ratio is high, equalling or surpassing unity in the early half of the emergence span. In Siteroptes cerealium, the usual male/female ratio is .04 to .05, in Pyemotes species, .035 to .0375 (after Herfs, 1926). The two sexes differ as greatly in their locomotory movements as in their physical appearance. Females exhibit a steady gait, usually utilizing only legs II and HI. Legs IV are occasionally used in walking, but generally seem to serve as organs of balance 6—1367 126 THe UNIVERSITY SCIENCE BULLETIN under conditions involving smooth terrain. The huge first pair of legs are used as sensory appendages in the same manner as those of Scutacaridae. They are carried slightly diverging before the advancing mite, often making alternate “pawing” motions in the air, or they may be tapped on the substrate as the mite proceeds. On several occasions the pawing motions were seen with legs I held vertically above the propodosoma. Males, on the other hand, are excessively awkward. Apparently, all four pairs of legs are used for locomotory purposes, but in an uncoordinated manner. Move- ments are halting and clumsy, often consisting only of slow, turn- ing motions, and they seem incapable of walking any great distance. No differences in gait were observed when males were isolated in separate dishes. (Perhaps the excessive awkardness of the male may help to explain the comparatively high male/female sex ratio in the species. ) Mating was observed in this species, the process differing con- siderably from that described for Siteroptes and Pyemotes. As soon as males emerge, they begin to wander about, during which time they may attempt to mate with any females encountered. It is probable that, as in Pyemotes, they are not stimulated by in- seminated females, since they make no attempt to copulate with many of the females passing within reach. Perhaps this may be a behavioral rather than a physiochemical response on the part of the female, since many females seem to actively avoid or rebuff the males. Having found a partner, the male slides underneath her body, still with his ventral surface toward the substrate. Reaching upwards with legs I and II, he grasps the anterior portion of the female hysterosoma and holds the female in this position during copulation, her venter pressed tightly against his dorsum. The apex of the male opisthosoma is then directed upward to make contact with the vulva of the female. Contact seems to last but an instant. In at least one instance, the female was newly hatched, having just previously shed her embryonic membrane. Males were never seen to pick up the quiescent nymphs and attempt copulation as has been described for related forms. Although certain parts of the life history seem clear, much is unknown and the following account is largely speculative. Young, presumably mated females are found within the nests of the host. Although they tend to congregate in masses within the cells in which they are born, this behavior must be later modi- THe Famity PYEMOTIDAE DY fied so that they escape from the cells and disperse inside the bee burrows. There they may find the original mother bee (in the case of Halictus farinosus), or the succeeding generation of bees emerging the same year, or (Nomia, Calliopsis) they overwinter within the burrow to find the brood emerging the following spring. In any event, it seems clear that mites leaving the nest usually are dispersed by the adult bees or their parasites or inquilines. Young female mites have been found clinging to the mesosoma or head of adult Halictus farinosus Smith, Nomia nevadensis bakeri (Cockerell), Sphecodes arvensiformis Cockerell, and an unidenti- fied mutillid. Concerning the latter two, the first is a known parasite of Halictus farinosus and it is likely that the second is also a bee parasite. In searching for a suitable host cell, the female S. arvensi- formis may puncture the caps of a sealed cell (G. Bohart and Cross, unpublished). It is possible that this behavior may facilitate the dispersal of mites from infested cells. The young females must obviously find a host larva. Those at- tached to female bees or bee parasites may drop off into cells newly constructed or provisioned and wait, probably inside the cell, for the bee larva to reach the proper stage of development. The author has seen slides of young female mites labelled “from the larva of Calliopsis andreniformis.” No cells found by the author in Utah contained both mites and bee larvae. Presumably the host is killed, perhaps by a toxic salivary secretion similar to that of Pyemotes sp. A. Shinn (in litt.), working with Calliopsis andreniformis in Kansas in 1957-58, reports the same finding. At least in H. farinosus the larva may be killed before it finishes feed- ing, as indicated by a small amount of pollen found in at least one parasitized cell of H. farinosus. It is presumed that the female feeds by first piercing the skin of the host with the needlelike chelicerae, then sucking body fluids. Of interest is the fact that both Shinn (in litt.) in Kansas and the author in Utah have never found more than two gravid female mites in a bee cell. As the female feeds, the entire hysterosoma begins to swell and become globelike, this continuing until she has increased her orig- inal size many times. The diameters of the hysterosomata of gravid specimens from Utah and from Kansas are 2.3 and 3.0 mm., re- spectively. Within this enormously distended sphere, the eggs are formed. 128 THE UNIVERSITY SCIENCE BULLETIN Resinacarus Vitzthum Resinacarus Vitzthum, 1927, Sitzungsberichte der Gesellschaft Naturforschender Freunde Berlin, 1927, pp. 89-110. Type, Resinacarus resinatus Vitzthum, 1927, by original designation. The single species resinatus included in this genus has not been seen since it was originally described. Resinacarus will therefore not be described herein and the reader is referred to the original description. Partly due to Kramer’s (1877b) incorrectly ascribing (see p. 156) but two pairs of dorsal propodosomals to Pygmephorus spinosus, confusion has existed in assigning forms to Resinacarus. Particularly in the United States, species properly belonging to Pygmephorus sensu stricto (as defined in this paper) have been incorrectly placed in this genus. In actuality, Resinacarus ap- pears to be a unique form and may be separated from the remaining genera in the family by the characters given in the above keys. The male of Resinacarus resembles that of some Pyemotinae and some Acarophenacinae in the following respects: functional gna- thosoma present; leg IV similar to that of the female; pronounced genital capsule lacking; propodosoma with four pairs of dorsal setae. The female resembles some Acarophenacinae as well as some Pyemotinae and differs from those of the Pygmephorinae in that coxa IV is triangular and similar to coxa III, because apodemes IV arise from the anterior sternocoxal condyles of legs IV, and in pos- sessing a pair of opisthosomal sternals; it resembles the Acaro- phenacinae and differs from the other two subgenera in having only three pairs of anterior sternal setae and four pairs of posterior ventral setae; it resembles some Pygmephorinae and Pyemotinae and differs from the Acarophenacinae in that coxa I appears to arise only slightly obliquely to the longitudinal axis of the body, in that the gnathosoma is elongate, free, and decumbent, and in body shape. In consideration of the above, the author tentatively places Resinacarus as a specialized genus (fused tibiotarsus I, loss of two pairs of ventral setae) within the subfamily Pyemotinae. Vitzthum took resinatus from tubes bored in spruce resin by Myolepta sp. and another unidentified species of syrphid fly. He believed the mite to be associated with the fungus Sporotrichum flavissimum, therefore similar to Siteroptes in food habits. THE FAMILY PYEMOTIDAE 129 Acarophenacinae, new subfamily DESCRIPTION OF NON-GRAVID FEMALE. Small mites, usually 200y. long or less, rarely 300y; shape oval to short-fusiform, posterior mar- gin of hysterosoma usually sharply rounded; integument weakly to moderately sclerotized, lacking “pebblings” or scalelike engravings; most lateral portions of posterior ventral plate not striate. Gnathosoma. Distinct and normally directed anteriorly or hidden within propodosoma; palps fused to gnathosoma, not distinct; chelicerae enlarged, distinct, needlelike or bladelike. Propodosoma. Dorsum. Entire or bipartite; pseudostigmata present or (usually) absent; two or three pairs of dorsal propodo- somal setae present; posterior margin of propodosoma meeting or overlapping anterior margin of first hysterosomal tergum; esophag- eal “finned structures” not visible, probably lacking. Venter. Anterior ventral plate with three pairs of setae or less; secondary transverse suture lacking. Hysterosoma. Three to five visible segments, segment V reduced, sometimes vestigial; segment V, if distinct, with two pairs of setae or fewer; posterior portion of vaginal wall not sclerotized or en- closed within enlarged, globose to elliptical pumplike structures; definitive anterior and posterior genital sclerites absent but various internal opisthosomal sclerites sometimes present; opisthosoma short, usually not much longer than distance between anterior mar- gin of coxa III and posterior margin of coxa IV; setae of opistho- somal venter present or absent. Legs. Leg I 4 or 5 segmented, arising at nearly a right angle to the longitudinal axis of the body or distinctly obliquely; trochanter I with three or four setae, seta c setiform; claws I and IV, II to IV, or I to IV may be absent or all claws may be present; pinnacula absent; coxa IV subcordate or bandlike, never quadrate or rec- tangular; tarsus IV not greatly elongate, never more than three and one-half times length of fermurogenu IV. Key to the Genera of the Subfamily Acarophenacinae (Females ) l(a). Leg I with five segments; pseudostigmata present....Caraboacarus, p. 130 1(b). Leg I with four segments; pseudostigmata absent................. 2 2(a). Leg IV with claws; peritremes opening dorsally; dorsal propodosomal U-shaped apodeme Nic ken ois ec rs eat Rr ae as, Paracarophenaz, p. 132 2(b). Leg IV without claws; peritremes opening anteriorly; dorsal, propo- dosomal U-shaped apodeme PRESSE AME ine en ene nnaieee era Neen ie 3 180 THe UNIVERSITY SCIENCE BULLETIN 3(a). Tibiotarsus I with stout, sessile claw; posterior ventral plate with five Pairs Of SEbAE L108 Whee erate) we rents aon yi. andr grace Ne? Acarophenax, p. 138 3(b). Tibiotarsus I clawless; posterior ventral plate with three or four pairs Of SCUBO: ciirane a) aw tae eke Abe eRe oo kya cat a ee Adactylidium, p. 142 Caraboacarus Krezal (Figs. 41-42) Caraboacarus Krezal, 1959, Beitrage zur Systematik und Okologie mitteleuro- paischer Acarina. Bd. I: Tyroglyphidae und Tarsonemini, Teil 2, p. 570. Type, Caraboacarus stammeri Krezal, 1959, by original designation. This genus is known from various carabid genera in Europe and the central United States. Although possessing several characters considered to be specialized, including the loss of claws I and IV, loss of the opisthosomal sternals, the fusion of tibia and tarsus IV, small size, the oblique articulation of coxae I, the possession of only two pairs of propodosomal setae, and the modified, paddlelike ventral setae, it is tentatively placed as the most primitive form in the subfamily because legs I are five-segmented, pseudostigmata are retained, the anterior ventral plate bears three pairs of setae, the posterior plate bears six pairs, gnathosoma is free, propodosomal dorsum is entire, trochanter I has four setae, tarsus II has seven setae, the solenidia of both tibiae II and III are present, and be- cause the fifth segment is relatively well developed. Despite its wide host range and distribution, the genus is known to include but two species, one undescribed. Only the non-gravid fe- males are known. Diagnosis. Females may be differentiated from those of all other genera by the following combination of char- acters: legs I five-segmented, gnathosoma free, at least twice as wide as long, palps fused, body short-fusiform, opisthosoma very short, projecting behind posterior margins of coxae IV only about as much as gnathosoma projects forward from anterior margins of coxae I. DeEscRIPTION OF NON-GRAVID FEMALE. Length, 160 to 173y.; width 111 to 120%; width between anterior condyles of coxae III 66 to 77y; hysterosoma ovate-fusiform; body weakly sclerotized or not, with or without coarse punctures. Gnathosoma. Distinct, bow-shaped, at least twice as wide as long, directed anteriorly; palp indistinct, fused with gnathosoma: chelicera stout, recurved, bladelike, easily visible; pharyngeal- esophageal pump present but dorsal rather than ventral. Propodosoma. Dorsum. Broadly trapezoidal, nearly rectangu- late; posterior margin meeting or hidden beneath first hysterosomal THe FamMiIty PYEMOTIDAE 131 segment (In mounted specimens of the same series, both conditions occur. Since the overlap between the two sclerites is small, these discrepancies probably occur during mounting); prominent dorsal U-shaped apodeme lacking; pseudostigmata present, normal, lo- cated in posterior one-half, arising well laterad; two pairs of propo- dosomal setae, one pair of pseudostigmatals missing, remaining pair less than twice length of stigmatals; peritermes circular to oval, situated at anterolateral corners of dorsum, often hidden or indis- tinct; two pairs of large, sclerotized “tubercles” present, one pair anteromesad of stigmatals, a second, larger pair behind pseudo- stigmatals; internal atria dorsal, elongate, moderately prominent. Venter. Ventrites I lamellate anteriorly or normal, sometimes (stammeri) with a pair of circular, suckerlike structures; anterior ventral plate with three pairs of setae, ventrites I with one pair, ventrites II with two pairs, external ventrals I lacking; line of articu- lation of coxa I oblique to longitudinal axis of body; apodemes II wide, becoming platelike laterally, arcuate, complete, making angle of about ninety degrees with anterior median apodeme, latter well developed, extending posteriorly to line drawn between posterior condyles of coxae II; pleated conjunctival membrane between propodosoma and hysterosoma not visible; posterior marginal apo- demes well developed medially, incomplete at sides. Hysterosoma. Dorsum. First segment entire, meeting or extend- ing slightly over posterior margin of propodosoma; fourth segment with two pairs of setae; segment V reduced, cuplike, concealed dorsally by fourth or nearly so, bearing one or two pairs of setae, inner pair tiny, spinelike, often indistinct; vulva apical; genital sclero- tizations variable, seemingly dependent upon mount. Venter. Posterior ventral plate entire, bearing six pairs of setae, at least one pair modified spinelike; second axillaries nearly directly mesad of anterior condyles of coxae IV; external presternals well laterad of and posterior to internal presternals, much closer to latter than to first axillaries; external poststernals arising directly behind but some distance from internal poststernals, near line drawn between most mesal margins of coxal foramina IV; apodemes II well developed, arcuate, incomplete medially; apodemes IV also well developed, straight, complete or nearly so, arising near posterior margins of coxae III and proceeding anteromedially at angle of about 45 de- grees; apodemes V absent; posterior median apodeme distinct, its posterior termination in vicinity of internal poststernals, band of 132 THe UNIVERSITY SCIENCE BULLETIN opisthosomal conjunctival pleats visible or not; opisthosoma char- acteristically short, much shorter than distance between anterior margins of coxae III and posterior margins of coxae IV; opisthoso- mal venter entire, one pair of setae present or not. Legs. Well developed, margins of segments often with well sclerotized areas of varying thickness; leg I with five segments, distinctly shorter and thinner than leg II; tarsus J short and taper- ing, narrower than other segments of leg; lateral sensory spot absent; claw I lacking; tibia about same breadth as femurogenu; trochanter I with four setae, seta c stout, long, aciculate, reaching nearly to largest tarsal solenidium; seta e peglike or clavate; trochanter II with three setae, femurogenu II with three; trochanter III with two setae, femurogenu III with three; tarsus II with seven setae and a sensory peg, tarsus III with seven setae; solenidia of tibiae II and III present, basal; claws II and III present, short and stout, with a large mesal lobe; claws IV lacking; pulvilli II and III much longer than claws; coxa III subcordate, coxa IV bandlike, distinctly mesad of coxa III; tibia and tarsus IV fused; tibiotarsal suture distinct half way around leg; most elongate seta of leg IV dilated bladderlike basally. Hasits. Because of the greatly enlarged chelicerae of these mites, it is probable that they are parasites. All specimens known to date have been taken from beneath the elytra of various carabid beetles, principally of the genera Harpalus (in Europe) and Agono- derus. SPECIMENS EXAMINED. Two species (one probably stammeri) from Kansas and Michigan, from the following hosts: Agonoderus comma Fabricius,® A. pallipes Fabricius, A. sp., “large black cara- bid,” “small black carabid.” ? Paracarophenax, new genus (Figs. 45-47) Type, Paracarophenax dybasi, new species. This genus includes at least four species, two of which are un- described. One of these, P. dybasi, is described herein. P. dermestidarum (Rack) and P. bambergensis (Krezal) are found upon various dermestids and upon nitidulids, respectively. The habitat of P. dybasi (in debris under decaying fruit) suggests a host similar to the latter. An undescribed species near P. dybasi 9. Det. G. E. Ball, 1960. THE FamMity PYEMOTIDAE 133 has been taken from a Panamanian cerambycid. Rack (1959), gives an excellent account of P. dermestidarum. Members of the group are closely related to Acarophenax but are considered less specialized in that claws I to IV are retained, in that the propodosomal arcula is less specialized, because segment V is relatively well developed and bears a pair of setae, and because the solenidium of tibia II is retained. In addition, the peritremes are dorsal in some (but not all) species. The male of only one species, P. dermestidarum, is known. Dracnosis. Females differ from those of all other genera in the family by the following combination of characters: anterior ventral plate with only two pairs of setae; gnathosoma concealed within the propodosoma or nearly so; chelicerae stout, easily visible; legs II to IV with claws. DEscRIPTION OF NON-GRAVID FEMALE. Length, 200 to 300u; width, 98 to 180y; width between anterior condyles of coxae III 65 to 110y; hysterosoma subovate, its posterior margin narrowly rounded; body weakly sclerotized and without coarse punctae. Gnathosoma. Outlines indistinct, nearly completely enclosed within propodosoma, not visible in dorsal aspect; palpi fused with gnathosoma, indistinct; chelicerae stout, styliform, recurved, easily visible; elongate, spindle-shaped pharyngeal-esophageal pump pres- ent, often indistinct, extending posteriorly as far as mesal portions of coxae II. Propodosoma. Dorsum. Dorsal propodosomal plate broad, shield- like, overlapping bases of legs II, projecting anteriorly, lobelike, be- tween legs I, sometimes ventrally to gnathosoma; posterior margin overlapping first hysterosomal segment; dorsal, U-shaped apodeme lacking; peritremes dorsal or dorsolateral, located near anterolateral extremities of propodsoma, not enclosing gnathosoma laterally; in- ternal atria drop-shaped, ill defined or not visible; pseudostigmata absent; two or three pairs of propodosomal setae, anterior pairs absent or distinct, if distinct then less than half as long or wide as remaining pairs, located laterad of cheliceral bases. Venter. An- terior ventral plate with two pairs of setae, ventrites I with one pair, second pair on or approximate to apodemes II; ventral ex- ternals I lacking; coxa I attached at a right angle to longitudinal axis of body or slightly oblique; apodemes II moderately to weakly developed, long, straight, incomplete medially, sometimes narrowly so, making angle of about 45 degrees to longitudinal plane of body; 134 THe UNIvEeRSITY SCIENCE BULLETIN anterior median apodeme absent to well developed; pleated con- junctival band between propodosoma and hysterosoma indistinct; posterior marginal apodemes ill defined. Hysterosoma. Dorsum. First dorsal segment entire, overlapped anteriorly by propodosoma; third segment with one or two pairs of setae; fifth segment reduced, cuplike, usually not completely hidden beneath fourth in dorsal aspect, bearing a single pair of apicoven- tral setae; vulva apical, apical portion of vagina not heavily sclero- tized or enclosed in heavily sclerotized pumplike structure. Venter. Posterior ventral plate with five pairs of setae, entire or with tri- partite posterior margin; second axillary well mesad of anterior sternocoxal condyle IV; one pair of poststernals lacking; apodemes III weak or lacking except at anterior sternocoxal condyles of coxae III; apodemes IV weakly to moderately developed, nearly straight, incomplete medially, arising from posterior margin of coxae III and proceeding anteromedially; apodemes V absent; band of pleated opisthosomal conjunctiva indistinct; opisthosoma short, distinctly shorter than distance between anterior margin of coxa III and posterior margin of coxa IV; opisthosomal venter entire, bearing a single pair of setae. Legs. Well developed and long; leg I four-segmented, distinctly stouter than remaining legs; tibiotarsus I subelliptical to rectangular, distinctly wider than other segments of leg; large, rounded, medio- lateral sclerotized area near outer margin of tibiotarsus I midway between base and claw; claw I single, stout, sessile, furnished with a large, toothed thumb; trochanter I with three setae, seta c stout and aciculate, not or scarcely reaching tibiotarsal sensory spot, seta d arcuate; trochanter II with three setae, femurogenu II with one or three; trochanter III with one or two setae, femurogenu III with one or three; solenidium of tibia II present, that of tibia III absent; tarsus II with six tactile setae and a sensory peg, tarsus III with six tactiles; solenidum of tarsus II in basal third of segment; pretarsi II to IV elongate, two-thirds to nearly as long as tarsi; claws II to IV well developed, simple, pulvilli II to IV also well developed; coxa III short-triangular in ventral aspect, coxa IV broadly cres- centic in ventral aspect, distinctly mesad of coxa III; femurogenu IV with one or two setae; tarsus IV only slightly longer than tibia IV, tapering evenly; ratio of femurogenu IV to tarsus IV about 1:2. DESCRIPTION OF GRAVID FEMALE.!° Gnathosoma as described for 10. Specimens of these stages not seen; information taken from original description of Acarophenax dermestidarum (Rack). THe FAMILY PYEMOTIDAE 135 non-gravid female; propodosoma and hysterosoma becoming phy- sogastric and spherical, sometimes reaching 700u.. DESCRIPTION OF Mate.!° Similar to that of Acarophenax; slightly smaller than female; posterior margin rounded; body dorsoventrally flattened; opisthosoma not greatly modified to form connate or trapezoidal genital capsule. Gnathosoma. Hidden within propodosoma, chelicerae functional. Hysterosoma. Venter. Generally agreeing with description of Acarophenax but appearing to possess two pairs of setae rather than two pairs of suckers on genital plate; apicoventral suckers also seem- ingly absent; apodemes III and IV are complete medially or nearly so. Legs. All legs shorter than those of female, leg IV subequal to or slightly longer than leg III, otherwise agreeing with the descrip- tion of Acarophenax. Species thought to belong in Paracarophenax: bambergensis (Krezal, 1959) ( Acarophenax ) dermestidarum (Rack, 1959) (Acarophenax ) dybasi, new species ipidarius (Redikortsev, 1947) ( Pediculoides ) Hasirs. Unknown except for the excellent account of the habits of P. dermestidarum (Rack). In general, the life history of that species agrees well with that of Acarophenax tribolii. Several in- teresting differences, and points not noted by Newstead & Duvall should perhaps be mentioned here, however. Females of P. dermestidarum differ from those of A. tribolii in their locations on the adult host, those of the latter occurring prin- cipally under the wings, those of the former preferring the venter, principally that of the thorax. Concerning the question of the young females of dermestidarum parasitizing adult beetles, see p. 141. Unlike those of A. tribolii, dermestidarum females will leave their adult host and actively seek food if the latter are incapable of oviposition. Rack mentions that complete physogastry may be attained following the consumption of about half the contents of a single beetle egg, the active feeding period lasting 24 to 30 hours. The number of offspring per female is higher than in tribolii, and ranges from 3 to 27. As in the latter species it is directly correlated with the amount of physogastry attained by the mother. It was ascertained by Rack that mating in P. dermestidarum occurs within the body of the mother, only a single male being 136 THE UNIVERSITY SCIENCE BULLETIN produced in any one female. After impregnating its sisters, this male often initiates the maternal hysterosomal breakdown and subsequent birth of the offspring by pushing its way through the body wall of its mother. If neighboring females are available, this newly escaped male may climb upon one of them and wait for mating to begin within. When such activity commences, it may then claw its way into the body of the mother concerned and assist the resident male in mating the females. During its period of waiting, this intruding male may occasionally puncture the body of the “host” mother to feed. The young females of dermestidarum, unlike those of tribolii, need no conditioning period upon the adult beetles before attack- ing their eggs to begin feeding and physogastry. SPECIMENS EXAMINED. In addition to the type series of dybasi, the following have been examined: Species No. 1. PANAMa: Barro Colorado Is., 29° @ from body of cerambycid. Paracarophenax dybasi,‘' new species Diacnosis. This species is closest to P. bambergensis (Krezal), from which it differs in that the posterior margin of the posterior ventral plate is distinctly tripartite, in that all dorsal setae except laterals IV are stout and sparsely plumose, and because tarsus II and tarsus III possess two and one spinelike setae, respectively. DESCRIPTION OF NON-GRAVID FEMALE. Length, 257y (257-279, X = 264.33); width, 172u. (172-182, X = 178.00); distance between anterior sternocoxal condyles III, 98y. (98-110, X = 105.00). Gnathosoma. Cheliceral bases widely separated, chelicerae large, bladelike, arcuate, converging posteriorly, then recurved abruptly anteriorly to enter buccal cavity. Propodosoma. Dorsum. Large, hemispherical except for deep anterolateral excavations marking insertions of coxae I; anterior mar- gin seemingly bending ventrally to cover gnathosoma anteriorly as well as dorsally; peritremes circular, dorsal, in anterior one-quarter, separated by less than twice the width of one; stigmatal openings circular, distinctly visible within peritremes; one pair of setae (stigmatals?) about one peritremal diameter in front of peritremal margin, short, bristlelike, directed anteriorly (position easily altered during mounting ); pseudostigmatals stout, spiculate, blunt apically, anterior pair subequal in length to posterior pair and arising less 11. This species is named in honor of Henry S. Dybas, Associate Curator of Insects, Chicago Natural History Museum, who collected the type series. THe Famity PYEMOTIDAE 1387 than two areolar widths anterior to a line drawn between them; anterior pseudostigmatals about 43» apart, posterior pseudostig- matals about 34y laterad of respective anterior pseudostigmatals. Venter. Setae of anterior ventral plate nude, short, ventrals II about 3ly. long, more than thrice length of ventrals I and distinctly mesad of them. _ Hysterosoma. Dorsum. All setae of dorsum except laterals IV stout, spiculate, apically blunt; laterals IV flagellate; dorsals I well posterior to laterals I, 2ly (21-24, X = 22.80) long, 68 (68-72, X = 69.40) apart; laterals I slightly longer than dorsals I, 26y. (25- 26, X = 25.60), closer to respective dorsals than dorsals are to each other; dorsals II as laterals I, 85y. (85-92, X = 90.00) apart; dorsals III 18 (15-19, X = 17.00), 15y (15-23, X = 20.60) apart, laterals III usually slightly longer than dorsals III and posterolaterad of them by about their own length; dorsals [TV 14y, (14-15) long, 20u. (20-24) apart; laterals IV 17p. (17-21, X = 18.30) long, well antero- laterad of dorsals IV; setae of segment V well ventral, short, about S to 9x. Venter. Posterior margin of posterior ventral plate shal- lowly tripartite; apodemes III vestigial, developed only at anterior sternocoxal condyles of coxae III; apodemes IV moderately de- veloped, weakly sinuate, terminating well anterior to but on a longi- tudinal line with internal poststernals; all setae of posterior ventral plate short, nude, flagellate, about 11 to 23y long; internal prester- nals arising in a straight line with external presternals and post- sternals; external poststernals arising from center portion of tri- partite ventral plate on (or nearly on) line drawn between posteromesal margins of coxal foramina IV; posterior margin of opisthosomal venter with distinct median excavation; opisthosomal ventrals 20y. (20-26, X = 22.80) apart, 23y (19-23, X = 21.20) long, nude, flagellate. Legs (all measurements in ».). Length. Leg I, 72 (69-72, KX = 69.60); leg II, 90 (90-94, X = 91.80); leg III, 96 (94-104, K= 97.40); leg IV, 110 (110-117, X = 112.00). Width. Leg I, 24 (22- 24, X= 23:60); leg Wy I (17-20) X = 18:21); leo Ii, 14 (14-15: X = 14.20); leg IV, 17 (15-17, X = 16.20). Segment lengths. Tr I, 22-( 22-24, X = 23:20); tita 1, 36 (34-36, X = 34:80); t& IL, 3h (@6- Slee xe—130-20)) ata WE oi (383-o,) & —34.60)- ti le NS (Gh ns: Me) ta WN CX Ba 00i) ex IN 1202932204 OX == 22'60)r tr IV, 42 (40-44, K = 42.40); fege IV, 13 (12-13, XK = 12.60); ta 12. Distance measured between centers of anteroventral and posterodorsal sternocoxal condyles. 138 THe UNiversiry SCIENCE BULLETIN IV, 48 (41-43, X = 42.40). Setation. Leg I: cx IT, t 31, fege AT, tita 14T + 5S; leg II: cx IT, tr 3T, fege 3T, ti 4T + 1S, ta (T+ 1S: leg MM: cx 1T, tr 20, fege 30, uw 40, tay i leo WWEnowiale tr 2T, fege 2T, ti 4T, ta 6T. Seta c of trochanter I stout, blunt apically, seta d recurved; tibiotarsus I with two clavate and three setiform solenidia; basal clavate solenidium distinctly longer than and half the width of the distal; first seta basad of proximal clavate solenidium setiform, not lanceolate or flagellate; claw I with distinct mesal tooth on in- ner surface, otherwise without rounded basal dilation; thumb pres- ent, enlarged, bidentate; trochanter Il, femurogenu II, and tibia II lacking spiniform setae, two such setae present on tarsus II, one on tarsus III; solenidium of tibia II elongate, weakly clavate, that of tarsus II less than half as long and twice as wide, both well distad of bases of their respective segments; tarsus IV lacking elongate, flagellate seta, slender, tapering evenly, ratio of femurogenu IV to tarsus IV 1.3.3 (1:3.2 to 1:3.6). Distribution. Known only from the type locality. Type MareriAL. Female holotype and four female paratypes (nos. 1-4) from S. W. shore, Lake Clay (nr. Lake Placid), High- lands Co., Florida, June 14, 1955, H. S. Dybas, in debris under de- caying fruit (berlese sample ). Type Reposrrories. Holotype and one paratype (no. 3) in the Chicago Natural History Museum, Chicago, Illinois. Two para- types (nos. 1, 4) in the Snow Entomological Museum, The Uni- versity of Kansas. One paratype (no. 2) in the U. S. National Museum, Washington, D. C. Acarophenax Newstead & Duvall (Figs. 43-44 ) Acarophenax Newstead & Duvall, 1918, Royal Society Report No. 2 of Grain Pests (War) Committee, pp. 19-25, pl. VII and VIII. Type, Acarophenax tribolii Newstead & Duvall, 1918, by original designation. This genus includes at least three forms, two of which are un- described. All are probably egg parasites of various beetles. The cosmopolitan A. tribolii attacks the eggs and perhaps the adults of Tribolium, but it has been recorded from other tenebrionids (Latheticus, Gnathocerus) as well. One undescribed form seen by the author was taken from the nest of a flicker (Colaptes), in which it was perhaps parasitic upon a resident beetle. A second (undescribed ) form is known from laboratory cultures of Crypto- THE Famity PYEMOTIDAE 139 lestes ferrugineus (Steph.). A. bambergensis Krezal is parasitic upon nitidulids. This last named species as well as A. dermesti- darum Rack, are placed in a new genus, Paracarophenax (p. 132). Dracnosis. The females differ from all other genera in the family by the following combination of characters: anterior ventral plate with two pairs of setae, gnathosoma reduced and fused with the propodosomal arcula, concealed or nearly so, chelicerae stout and easily visible, legs II to IV lacking claws and possessing pretarsi which are nearly as long as the tarsi. DEscCRIPTION OF NON-GRAvVID FEMALE. Length, 105 to 198u; width, 66 to 128y, width between anterior sternocoxal condyles of coxae III, 40 to 83y; hysterosoma narrowly to broadly oval, posterior margin narrowly rounded; body weakly sclerotized, without coarse punctures. Gnathosoma. Outlines indistinct, seemingly at least partially fused with subdorsal propodosomal arcula, not or scarcely visible in dorsal aspect; palpi fused, indistinct; setation reduced, consisting of one pair anteriorly or anterodorsally, one pair of ventral palpal setae and sometimes a pair of palpal solenidia, and one or two pairs of lateroventrals, the most basal of which may be the displaced internal ventrals I; chelicerae stout, styliform, recurved basally, easily visible; large spindle-shaped pharyngeal-esophageal pump visible internally, extending posteriorly to anterior mesal portions of coxae II, salivary ducts distinct dorsal to this pump. Propodosoma. Dorsum. Dorsal plate broad, shieldlike, at least partially projecting over gnathosomal region anteriorly, posterior margin overlapping first hysterosomal segment; internal U-shaped apodeme of arcula prominent dorsally, its arms continuous anteriorly with peritremes; peritremes in “shoulders,” directed anteriorly or turning ventrally, enclosing gnathosoma laterally; internal atria prominent, dorsal, drop-shaped or circular; pseudostigmata lack- ing; two or three pairs of propodosomal setae, if three, anterior pair small, indistinct. Venter. Anterior ventral plate with two pairs of setae, ventrites I and II each with a single pair; ventral externals I lacking; coxa I (as seen in ventral aspect) attached at a right angle to longitudinal axis of body or nearly so; apodemes II very strong, straight, incomplete medially or nearly so, making angle of about 45 degrees or less with anterior median apodeme, the latter well developed; posterior marginal apodemes well defined, nearly complete; pleated conjunctival band between propodosoma and hysterosoma indistinct. 140 THE UNIVERSITY SCIENCE BULLETIN Hysterosoma. Dorsum, First segment entire, overlapped ante- riorly by propodosoma, fourth segment with one (rarely) or two pairs of setae, completely enclosing fifth segment in dorsal aspect; fifth segment much reduced, sometimes vestigial, bearing one pair of setae or none; vulva apical or apicoventral; internal hysterosomal structures absent or at least indistinct. Venter. Posterior ventral plate entire, bearing three or five pairs of setae; hind margin of posterior ventral plate broadly rounded or shallowly trilobate; second axillaries well mesad of and slightly anterior to legs IV; external presternals (may be internal poststernals) lacking; external poststernals arising medially or absent; apodemes III rather poorly developed, absent or short and straight and directed anteromedially; apodemes IV strong, straight, incomplete, arising at anterior sterno- coxal condyles of coxae IV and proceeding anteromedially; apo- demes V lacking; posterior median apodeme absent or reduced to a short line posterior to mesal terminations of apodemes IV; band cf opisthosomal conjunctival pleats indistinct; opisthosoma short, distinctly shorter than distance between anterior margin of coxa III and posterior margin of coxa IV; opisthosomal venter entire, bearing a single pair of setae. Legs. Well developed and long, margins of segments often with well sclerotized areas of varying thickness; leg I four-segmented, distinctly wider than leg II and (excluding pretarsus ) about as long; tibiotarsus I subelliptical, distinctly wider than other segments of leg; large, rounded sclerotized spot near outer margin of tibiotarsus I present midway between base and claw, this spot appearing in- ternal rather than on dorsal surface; claw I single, stout, dilated basally, sessile, sometimes deformed, furnished with a stout, spine- like thumb, or thumb vestigial; trochanter I with two or three setae, seta c aciculate or absent, if present not reaching tibiotarsal sensory spot; trochanter II with three setae, femurogenu II with one or three; trochanter III with one or two setae, femurogenu III with one or two; tarsus II with five or six setae and a sensory peg, tarsus III with five setae; solenidium of tarsus II in basal one-half; solenidia of tibiae II and III lacking; pretarsi of legs II to IV elongate, nearly as long as tarsi; claws II to IV lacking, pulvilli II to IV well developed; coxa III long, triangular; coxa IV subcordate, distinctly mesad of coxa III; femurogenu IV with one seta, tarsus IV moderately long, tapering evenly, ratio of femur eae IV to tarsus IV approximately 1:2.2 to 1:3.1. THE Famity PYEMOTIDAE 141 DESCRIPTION OF GRAVID FEMALE.'® Gnathosoma as described for non-gravid female but entire hysterosoma becoming greatly phy- sogastric, globular. DEscRIPTION OF MateE.'* Length, similar to female; more angulate posteriorly than female and flattened dorsoventrally in life; opis- thosoma not greatly modified to form connate or trapezoidal copula- tory structure. Gnathosoma. Reduced but present, hidden within propodosoma. Hysterosoma. Venter. Apodemes III, IV, and V well developed, the first two nearly complete medially; opisthosoma bearing large, subcircular “genital plate,” this with two pairs of “minute suckers”; apicoventrally, a pair of long setae and “another pair of suckers.” Legs. All legs five-segmented, legs II to IV shorter and stouter than those of female; tarsus I less incrassate, claw I smaller than in female; legs IV not tonglike or greatly modified for clasping, similar to legs III but distinctly shorter, Hasits. Information is available only on the habits of A. tribolii. This species is found upon the bodies of Tribolium confusum Duval and other grain-inhabiting forms, principally beneath the elytra. In beetle cultures nearly every specimen may become infested with as many as thirty but more commonly ten to twelve female mites. Newstead and Duvall state that they “appear to be able to attack their hosts where the cuticle is thin between the joints, but most of them . . . where the large area of soft cuticle . . . of- fers little resistance to their minute, piercing and sucking jaws.” In referring to a closely related form, Paracarophenax dermestidarum, Rack (1959), states that a similar mite-adult beetle association is purely phoretic. Since Newstead and Duvall failed to find gravid females upon the adult beetles, and since they report that the pres- ence of the mites did not harm the adult host, it is probable that the same situation obtains in A. tribolii. Phoretic females drop from the host beetle during the latter’s ovi- position and seek out an egg. Attaching to it, they begin to suck out its contents, resulting in its collapse. As they feed, physogastry commences, the entire female hysterosoma rapidly becoming en- larged and spherical, finally reaching 250 to 300y in diameter. As enlargement progresses, the capability of the female to move is lost. Newstead and Duvall, referring to the female, state that “on the 13. Specimens of these stages not seen, information taken from original description of A. tribolii N. and D. Quotations are from original description to which the reader is referred for more detail. 142 THe UNIVERSITY SCIENCE BULLETIN third day (after attachment to the egg) she dies and the young females emerge through a hole which corresponds in position to the genital orifice but which they enlarge for themselves.” As in Pyemotes and Trochometridium the definitive immature stages are thus lost and the preimaginal period greatly shortened. The num- ber of young varies from four to fifteen, depending upon the size of the physogastric mother. One of these is usually a male. New- stead and Duvall report several instances in which a male was lack- ing but never found more than a single male. Presumably mating occurs within the body of the mother before hysterosomal break- down and accompanying birth since the male is found (usually dead) only within or immediately outside of the maternal body. It was found that the newly born females did not immediately parasitize eggs offered to them, a fact which led Newstead and Duvall to infer that a conditioning period upon the adult beetle is necessary for “proper development.’ The fate of young females at- taching themselves to males or to very old female beetles is not known, although Newstead and Duvall report that many finally perish upon the beetles. SPECIMENS EXXAMINED. ‘Three species, including tribolii, from many parts of the United States, and India, taken from the follow- ing situations and hosts: rye mill, waste mill material, debris in nest of Colaptes cafer (Gmelin) (Aves, Piciformes), Tribolium confusum Duval, Gnathocerus cornutus (Fabricius), Latheticus oryzae Waterh., Cryptolestes ferrugineus (Steph. ), (all Coleoptera). Adactylidium, new genus (Figs. 48-50 ) Type, Adactylidium beeri, new species. This genus includes A. nicolae (Krezal) and three previously undescribed species taken from various Thysanoptera in Michigan, New York, Florida, and Colombia (S. A.). Morphologically, it is most closely related to Acarophenax, but differs in that it lacks claws I, in that apodemes II are weakly arcuate and join the anterior median apodeme at an angle of nearly 90 degrees, because apo- demes III are complete medially, and in exhibiting a reduced num- ber of setae on certain structures (anterior ventral plate with one or two pairs, posterior ventral plate with only four pairs, tarsus II with five setae plus solenidium, tarsus III with four). Of all the Acarophenacinae, members of this genus are alone in THE Famity PYEMOTIDAE 143 their association with insects other than beetles. As with other members of the subfamily, the species seem to be parasites of their hosts. Only the young, non-gravid females are known. Diacnosis. Females are distinct from those of all other genera in the family in possessing the following combination of characters: small size (length, 140 to 170p.), oval; leg I four-segmented, claws lacking on all legs; anterior ventral plate with one or two pairs of setae; pseudostigmata lacking. DESCRIPTION OF NON-GRAVID FEMALE. Length, 145 to 170yu.; width, approximately 90 to 110y; width between anterior sternocoxal con- dyles of coxa III approximately 44 to 68y; hysterosoma oval, its posterior margin usually narrowly rounded; dorsum with or with- out coarse punctae. Gnathosoma. Outlines indistinct, completely enclosed within propodosoma, not visible in dorsal aspect; palps fused, indistinct; setae of gnathosomal region reduced to one pair or none, if present, then directed anteriorly or anteroventrally; chelicera recurved, stout and needlelike, or stout basally and becoming indistinct apically; elongate, spindle-shaped pharyngeal-esophageal pump distinct or not, if distinct then extending posteriorly to anterior-mesal portions of coxa II, salivary ducts usually distinct dorsal to this pump. Propodosoma. Dorsum. Broadly trapezoidal, anterior margin in- distinct but probably continuing ventrad to completely or nearly completely envelop gnathosoma anteriorly; posterior margin over- lapping first hysterosomal segment; U-shaped apodemes of arcula prominent, continuous anteriorly with peritremes; peritremes in “shoulders,” circular, directed anteriorly or anteroventrally, bound- ing gnathosoma laterally; internal atria usually prominent, drop- shaped; pseudostigmata lacking; two or three pairs of propodosomal setae, anterior pair, if present, tiny, often indistinct, arising mesad of peritremes and appearing to be on gnathosoma. Venter. Ante- rior ventral plate probably with two pairs of setae, anterior pair sometimes not visible; coxa I as seen in ventral aspect attached obliquely to longitudinal axis of body; apodemes II strong, arcuate, complete medially or nearly so, making nearly a right angle with anterior median apodeme at point of intersection; anterior median apodeme well developed, its anterior portion appearing to extend into gnathosomal region; posterior marginal apodemes well defined, incomplete medially; pleated conjunctival band present dorsally between propodosoma and hysterosoma. 144 THe UNIVERSITY SCIENCE BULLETIN Hysterosoma. Dorsum. First segment entire, overlapped ante- riorly by propodosoma, bearing two pairs of setae; fourth segment with two pairs of setae, fifth segment much reduced, lacking setae, completely enclosed dorsally and laterally by fourth, usually visible only ventrally as an arcuate line; vulva apical. Venter. Posterior ventral plate entire, bearing four pairs of setae, external presternals and one pair of poststernals missing; hind margin of posterior ventral plate straight (could not be determined for one species); first axillary moderately mesad of anterior condyles of third coxa; second axillary seta arising from or slightly behind apodeme IV, well mesad of anterior sternocoxal condyle III; external presternal and internal poststernal setae lacking; apodemes III well developed, complete, arcuate, curving into posterior median apodeme; apo- demes IV strong, slightly arcuate, incomplete medially, arising at anterior condyles of coxae IV, in general paralleling apodemes III; apodemes V absent, an integumental fold between the posteromesal margins of coxae IV sometimes distinct and appearing apodeme- like; posterior median apodeme weakly to strongly developed, com- plete or intermittent, extending well behind line drawn between anterior sternocoxal condyles of coxae IV; band of opisthosomal conjunctival pleats well developed, usually distinctly visible be- neath cpisthosomal venter; opisthosoma short, distinctly shorter than distance between anterior margin of coxa III and posterior margin of coxa IV; opisthosomal venter bearing a single pair of setae, its posterior margin entire or deeply cleft. Legs. Well developed, of moderate length, margins of segments often with well sclerotized area of varying thickness; leg I four- segmented, distinctly narrower than leg II; tibiotarsus I elongate, sinuate or subrectangulate, not wider than other segments of leg; mediolateral sclerotized area near outer margin of tibiotarsus I lacking; claw I lacking; most basal seta of tibiotarsus I short, modi- fied bladelike or clavate; trochanter J with three small setae, seta c unmodified, not reaching base of tibiotarsus I; trochanter II with three setae, femurogenu II with three; trochanter III with two setae, ftemurogenu III with two; solenidia of tibiae IT and III lack- ing; tarsus II with five setae and a solenidium; tarsus III with four setae (one of these setae on tarsus II and two setae on tarsus IV arising at base of pretarsus, difficult to see); pretarsi II to IV moderately long, excluding pulvilli about half as long as tarsi; claws II to IV lacking, pulvilli well developed; coxa III long-triangular; coxa IV subcordate or short-triangular, not distinctly mesad of THe Famity PYEMOTIDAE 145 coxa III; femurogenu IV with one seta; tarsus IV about as long as III, tapering evenly; ratio of femurogenu IV to tarsus IV about 1:2.5 to 1:3.0; tarsus IV bearing one greatly elongate flagellate seta. Hasitrs. No information concerning the habits of these forms is known. On morphological and phylogenetic grounds one may infer that they are parasites, probably either of the eggs or other stages, of various species of thrips, and that the females become greatly physogastric when gravid. SPECIMENS EXAMINED. In addition to the types species, the fol- lowing are known: Species No. 1; Cotompta: Intercepted at Wash- ington, D. C.,5 9 ¢ from thrips on Cattleya sp. Species No. 2; Fioria: Welaka, 1 @ on Elaphrothrips tuberculatus (Hood);™ New York: Ithaca,3¢ 9? on Megalothrips spinosus 14 Hood. Adactylidium beeri,!’ new species Diacnosis. Very close to A. nicolae (Krezal), differing chiefly in that the single clavate solenidium of tibiotarsus I is in the apical half of the segment, in that the anterior setae of the anterior ventral plate (see Krezal, 1959b, p. 554) are not visible and perhaps lack- ing, and because the setae of segment III are subequally spaced. DeEscRIPTION OF NON-GRAvVID FEMALE. Length, 129. (128-142, X = 132.16); width, 89y. (89-96); distance between anterior sterno- coxal condyles of coxae III, 44y. (44-46, X = 45.00); body broadly oval. Gnathosoma. Chelicerae stoutly aciculate, converging posteriorly, then recurved abruptly anteriorly to enter buccal cavity; anterior setae of gnathosomal region lacking. Propodosoma. Dorsum. Dorsal propodosomal plate not coarsely punctate, extending posteriorly distinctly beyond line drawn be- tween posterior margins of coxal foramina II; both pairs of pseudo- stigmatals nude, finely flagellate, the inner pair approximately 15v. anteromesad of the lateral pair; distance between inner pseudostig- matals 37y. (37-39); internal U-shaped apodemes of arcula bandlike; peritremes opening anterolaterally, not or only slightly directed ventrad. Venter. Apodemes II rather evenly and shallowly arcuate, complete medially; ventral setae I not visible, ventrals I short, nude, flagellate. 14. Det. J. D. Hood, Cornell University, Ithaca, N. Y. 15. Named for Dr. Robert E. Beer, The University of Kansas. 146 THE UNIVERSITY SCIENCE BULLETIN Hysterosoma. Dorsum. Integument not coarsely punctate; setae of terga I to II long, nude, finely flagellate,!° setae of tergum IV short, bladelike; dorsals I less than diameter of areolus anterior to laterals, approximately 16 to 2ly long, 37p (37-43, X = 40.80) apart; laterals I about as long as dorsals I, nearly twice as close to respective dorsals as dorsals to each other; dorsals II subequal to dorsals I in length, 52y. (48-54, X = 52.00) apart; all setae of third tergum similar to those of dorsals I and II in length, dorsals II 17y. (17-22, X = 19.00) apart, laterals about two areolar diameters behind line drawn through their areoli, and well laterad of them; setae of tergum IV terminal, arising posteroventrally, directed pos- teriorly and posterolaterally, dorsals and laterals subequal in length and shape, seven to eight ». long, spaced subequally across segment or nearly so. Venter. Apodemes III stout, broadly arcuate; all sternal setae short, approximately eight to fourteen yp, nude, flagel- late; posterior margin of posterior ventral plate not or shallowly emarginate; internal presternals arising approximately eight y. lat- erad of posterior median apodeme, poststernals slightly more mesad; opisthosomal sternals nude, flagellate, short, approximately seven to eleven », on longitudinal line with internal presternals; posterior margin of opisthosomal sternum deeply emarginate. Legs (all measurements in y.). Length. Leg I, 33 (32-34, KX = 33.00); leg II, 46 (46-47, X = 46.33); leg III, 50 (47-52, X = 49.66); leg IV, 46 (46-52, K = 48.50). Width. Leg I, 8 (7-8, X =7.83); leg II, 13 (10-13, K = 11.66); leg III, 9 (8-9, K = 8.83) les Iv, 8 (8-9, K=8.16). Segment lengths. Tr 1, 9 (8-10; X = 9100): tita I, 19 (18-20, KX = 18.33); tr I, 14 (14-15, X= 14.83) eetae le 18 (15-19; X = 17.00); ti TM, 9 (X= 9.00), ta IN 197s xe 18.50); ex IV,17 14 (13-17, X = 14.50); tr IV, 15 (15-19, K = 16.66); fege IV, 5 (5-6, X = 5.20); ta IV 18 (18-19, X = 18.83). Setation. Leg I: ex OT, tr 31, fege 47, tita 117 - 5S; leg I: vex Oia fege 8T, ti 4T, ta 5T + IS; leg III: cx OT, tr QT, fege IT, ti 4T, ta AT; leg IV: cx OT, tr IT, fege 1T, ti 4T, ta (? 3T).18 Setae c and d of trochanter I short, setiform; tibiotarsus I subellip- tical, not sinuate, its most basal seta short (often difficult to see) with one median clavate and four apical-subapical rodlike solenidia; 16. Because of the extreme fineness of the apices of these setae, it was not possible to determine their lengths with enough accuracy to give exact measurements. Approxima- tions of their lengths are given instead. 17. Measured between anterior and posterior sternocoxal condyles. 18. Presence of apical setae not ascertained. THE FAMILY PYEMOTIDAE 147 solenidium of tarsus II short and rodlike, as broad as base of broad- est sensory seta of tarsus II. DistRIBUTION. Known only from the type locality. Type Marerrat. Female holotype and five female paratypes (nos. 1 to 5) from the University of Michigan Biological Station, Cheboygan Co., Michigan, August 2, 1958, R. E. Beer, from thorax of Thysanoptera (host No. M-58-5). Type ReposrroriEs. Holotype and three paratypes (nos. | to 3) in the Snow Entomological Museum at The University of Kansas; paratype no. 4 in the U.S. National Museum; paratype no. 5 to the Zoologisches Institut, Universitat Erlangen, Erlangen, Germany. Pygmephorinae, new subfamily DESCRIPTION OF NON-GRAVID FEMALE. Small to moderately sized mites, 150 to 600v. long; shape variable, usually elliptical to oval, rarely circular; posterior margin of hysterosoma evenly rounded or truncate; integument usually moderately, often heavily sclerotized, integumental “pebblings” or other engravings present or absent. Gnathosoma. Always free, directed anteroventrally, ventrally, or posteroventrally, never largely or completely hidden within the propodosoma; palps free and movable; chelicerae usually small, indistinct, rarely enlarged and distinct. Propodosoma, Dorsum. Always entire, pseudostigmata present; one to three pairs of dorsal propodosomal setae; anterior margin of first hysterosomal tergum usually overlapping propodosoma to varying degrees, occasionally entirely; esophageal “finned structures” present or absent. Venter. Anterior ventral plate with four to six pairs of setae. Hysterosoma. Four to five visible segments, segment V often en- closed dorsally but never posteroventrally by tergum IV; segment V with two or three pairs of setae; posterior portion of vaginal wall sclerotized or enclosed in enlarged, elliptical to globose pumplike structures or not; at least one genital sclerite present, shape variable; opisthosoma not distinctly shorter than distance between anterior margin of coxa III and posterior margin of coxa IV, or if so, then body circular; opisthosomal venter lacking setae. Legs. Leg I four-segmented; coxa I, as seen from beneath, articulated at a right angle to longitudinal axis of body or (usually ) distinctly obliquely; claws sometimes absent on leg I or leg IV, or 148 THe UNIVERSITY SCIENCE BULLETIN both, always present on II and III; trochanter I with three to four setae, seta c setiform, stout and rodlike, spatulate apically, or mem- branous and blade or hooklike; pinnacula present or absent; coxa IV quadrate to rectangulate. Key to the Tribes of the Subfamily Pygmephorinae (Females) 1(a). With three pairs of propodosomal dorsals; four to six pairs of anterior sternals; trochanter J with three or four setae; claw I, if present, sessile Pygmephorini, new tribe, p. 148 1(b). With one or two pairs of propodosomal dorsals; four pairs of anterior sternals; trochanter I with three setae; claw I, if present, usually more or less ‘pedicellate 2. 3. beh ls hs en, 2 2(a). Seta c of trochanter I setiform; one pair of propodosomal dorsals, or, if two, second pair not visible except under very high magnification; hys- terosomal segment V usually with two pairs of setae Microdispini, new tribe, p. 164 2(b). Seta c of trochanter I semimembranous, bladelike, the apical portion usually proflexed; one (rarely) or two pairs of propodosomal dorsals; segment V usually with three pairs of setae Neopygmephorini, new tribe, p. 196 Pygmephorini, new tribe This tribe includes three genera: Pediculaster Vitzthum, 1931, Pygmephorus Kramer, 1877, and Microdispodides Vitzthum, 1914. Its members are considered to be closely related to and perhaps directly descended from members of the genus Siteroptes. Females are easily distinguished from those of the remaining two tribes of the Pygmephorinae in possessing three pairs of propodosomal dorsals as well as in other characters outlined in the description below. DESCRIPTION OF NON-GRAVID FEMALE. Striae absent on most lateral portions of posterior ventral plate. Gnathosoma. Quadrate or longer than wide; palps never greatly elongate, fitted for piercing. Propodosoma. Dorsum. With three pairs of setae; peritremes dorsal, contiguous or widely separated, cylindrical, guttate, or circular; two esophageal “finned” structures present. Venter, Ante- rior ventral plate with four to (usually) six pairs of setae; external ventrals I furcate or not. Hysterosoma. Dorsum. First hysterosomal tergum never cover- ing more than posterior half of propodosoma; segment V with three pairs of setae; posterior portion of vaginal wall often enclosed in elliptical or globose pumplike structure; posterior genital sclerite usually elongate, never broadly triangular or flaplike. Venter. Apodemes V present or absent. THe Famity PYEMOTIDAE 149 Legs. Leg I articulated obliquely or at a right angle to the longitudinal axis of body; size variable; claw I rarely absent, if present, sessile; mediolateral sclerotized area of tibiotarsus I present or absent; trochanter I with three or (usually) four setae, seta c short, membranous, bladelike, often hooked, or elongate, stout- setiform, blunt or spatulate apically; pinnacula of tibiotarsus I absent. Key to the Genera of the Tribe Pygmephorini (Females ) 1(a). Seta c of trochanter I straight and stout, obtuse or spatulate apically Pediculaster, p. 149 1(6). Seta c of trochanter I short, semimembranous, more or less bladelike andspronexedvapicallyg ewes ae ee eed eam irene Monae nena ne eredegs 2 2(a). Claws I and IV lacking; dorsum of body with scalelike integumental ENG TAWA See rs OR as yates i cenica a ae a Microdispodides, p. 159 2(b). Claws I and IV present; dorsum of body lacking scalelike integumental CNET AV AID Sie eee mites eam ent Seana ie ine tee deere Pygmephorus, p. 156 Pediculaster Vitzthum This genus is closely related to certain Siteroptes and Pygme- phorus. It resembles the former in general body conformation, in that coxae J arise nearly at a right angle to the longitudinal axis of the bedy, in possessing three pairs of propodosomal dorsals, in that trochanter J bears four setae, and in that the stigmata are guttate or elliptical and contiguous or nearly so. It differs from Siteroptes and resembles Pygmephorus in that tibia and tarsus I are fused, in that claw I is stout and sessile, and because claws II and III are cleft and dilated basally. It also differs from Siteroptes in that many, if not all, forms are phoretic upon various Diptera. Specific relationships within the group are seemingly complex. The author has examined at least seven species. Only the females of Pediculaster are known. Dracnosis. Females are distinguished from those of other genera in the family by the following combination of characters: leg I with four segments, claw I sessile and often deformed, seta c of trochanter I stout and straight, blunt or spatulate apically, propodosoma with three pairs of dorsals. DEscrIPTION OF Non-cRAvID FEMALE. Length, 200 to 480y; width, 95 to 290; width between anterior sternocoxal condyles of leg III 65 to 240u; hysterosoma broadly to narrowly elliptical; integument pale and weakly sclerotized to yellow-brown and heavily sclerotized, lacking scalelike or “pebbled” integumental engravings; body usually with large punctae. 150 THE UNIVERSITY SCIENCE BULLETIN Gnathosoma. Normally directed ventrally, shape quadrate to rectangular; palps with large apical teeth; chelicerae distinct, acicu- late; esopbageal-pharyngeal pump absent but esophageal “finned” structures usually distinct. Propodosoma. Dorsum. Distinct dorsal sclerite present, en- larged posteriorly, rounded or falcate at points of greatest width; anterior portion prolonged hoodlike over gnathosoma, circumgna- thosomal foramen anteroventrad, its lateral margins only moderately sclerotized; posterior margin covered by anterior margin of first hysterosomal tergum or these two sclerites meeting but not over- lapping; pseudostigmata located in posterior third to fifth of dorsal plate; three pairs of propodosomal dorsals; peritremes dorsal and median, usually contiguous or nearly so, diverging, guttate, cam- panulate, or elliptical, often septate; two esophageal “finned” struc- tures visible. Venter. Anterior ventral plate with four to (usually) six pairs of setae, if five pairs then ventrites II with only two pairs; external ventrals I not bifurcate; ventral medials II, if present, usually much closer to externals II than to internals II; coxa I articulating at nearly a right angle to the longitudinal axis of body or (rarely) coxa I articulating obliquely; apodemes II distinct, straight or bending posteriorly near anterior median suture, usually complete, if straight then making angle of more than 45 degrees with anterior median apodeme; secondary transverse suture ab- sent; posterior marginal apodemes well developed; pleated con- junctival band between propodosoma and hysterosoma not distinct. Hysterosoma. Dorsum. Fifth dorsal segment well developed but usually completely enclosed by fourth in dorsal aspect; vulva apicoventral; vagina heavily sclerotized apically or not, surrounded by globose to elliptical pumplike enlargements or not; anterior genital sclerite separable from other genital sclerites or not, but if so, then its anteriormost point well posterior to line drawn between the posterior margins of coxae IV; posterior genital sclerite variable in shape but not broadly flaplike. Venter. Posterior ventral plate entire or (rarely) divided by a single transverse line between the centers of coxae III, the two plates thus formed contiguous and not overlapping; posterior ventral plate with four to six pairs of setae; hind margin of pesterior ventral plate tripartite, trilobed, or straight; second axillaries present or absent; external presternals well behind and usually barely laterad of internal presternals, aris- ing from or only slightly anterior to apodemes IV; position of ex- ternal poststernals variable; apodemes III and IV well developed THE Famity PYEMOTIDAE IES or (rarely) lacking, if present at least apodemes III meeting pos- terior median apodeme; apodemes III bending posteriorly at the sides to join anterior sternocoxal condyles III; apodemes IV arising laterally between coxae HI and IV, meeting posterior sternocoxal condyle of III; apodemes V present or absent, if present then join- ing posterior median apodeme; band of distinct opisthosomal con- junctival pleats lacking; opisthosoma long, always longer than dis- tance between anterior margin of coxa III and posterior margin of coxa IV. Legs. All legs well developed and bearing claws; leg I four- segmented, always distinctly wider and usually much longer than leg I; tibiotarsus I quadrate to rectangulate, tapering, not or barely wider than other segments of leg; pinnacula lacking; mediolateral sclerotized plate of tibiotarsus I absent or poorly developed; claw I stout, sessile, dilated basally or not, bearing thumb or not; tro- chanter I with four setae; seta c straight and stout, nude, not flagellate, bluntly pointed or spatulate apically, rarely reaching mid- dle of femurogenu; tarsi II and III each bearing six tactiles; soleni- dia of tibiae II and III present; claws II and II simple or with cleft or basal dilation; claw IV simple, usually not much smaller than those of II and III; pulvilli II to IV trumpet-shaped, always longer than claw; coxa III triangular or subcordate in inner ventral aspect; coxa IV subquadrate to rectangulate, not constricted near base, not bulbose; tarsus IV short to moderately elongate, not evenly attenuate from base to apex; ratio of femurogenu IV to tarsus IV approximately 1:1.6 to 1:4.1. Description of gravid female. As above but entire hysterosoma becoming physogastric, globular. Key to the Subgenera of Pediculaster (Females ) 1(@). Posterior ventral plate entire....2-.-..:. Pediculaster sensu stricto, p. 154 (b). Posterior ventral plate divided transversely between coxae III Acarothorectes, p. 151 Acarothorectes, new subgenus (Figs, 54-56) Type, Pediculaster ( Acarothorectes ) curculionium, new species. This subgenus is established to accommodate a single species which differs markedly from the remaining forms within the genus. In addition to the morphological differences mentioned below, the type series was taken from the body of a curculionid, an unusual host for forms of Pedicaster. 152 THE UNIVERSITY SCIENCE BULLETIN DeEscRIPTION OF NON-GRAVID FEMALE. Agrees with the descrip- tion of the genus except for the following: hysterosoma narrowly elliptical, markedly elongate; integument moderately sclerotized, coarsely punctate. Gnathosoma, Small, rectangulate. Propodosoma, Dorsum. Dorsal sclerite elongate, with distinct median longitudinal keel from which arise the peritremes, pseu- dostigmata, stigmatal, and anterior pseudostigmatal setae; laterad of kee! on each side, a concavity which receives the clavola of the pseudostigmata; peritremes campanulate, diverging nearly 180 de- grees; esophageal “finned” structures characteristic, the anterior much larger and more conspicuous than the posterior, the fins much more distinct and thicker than those found in other members of the genus. Venter, Anterior ventral plate with five pairs of setae; coxa I articulating at nearly a right angle to longitudinal axis of body; apodemes II straight, incomplete medially; anterior median apodeme complete, becoming increasingly weaker posteriorly. Hysterosoma. Dorsum. Vagina not heavily sclerotized apically, not surrounded by globose to elliptical pumplike enlargement, but a heavily sclerotized ringlike structure present. Venter. Posterior ventral plate divided by a single transverse line between the cen- ters of coxae III, the two plates thus formed contiguous and not overlapping; posterior ventral plate with but four pairs of setae; hind margin of ventral plate with single, large median lobe; internal presternals displaced anteriorly, arising at least two setal lengths anterior to external presternals, well mesad of latter; external post- sternals arising less than three areolar widths behind line drawn between posterior margins of coxal foramina IV; apodemes III to V absent; posterior median apodeme short, weak. Legs. Leg I much wider and distinctly shorter than leg II, tibio- tarsus I quadrate, tapering, not wider than femurogenu I; medio- lateral sclerotized plate absent; claw I with irregular basal dilation, thumb small; seta c spatulate apically;'® claws II and III with elongate ventral thickenings which terminate subapically, possibly in a tooth; claws IV simple, stouter than claws II and III; pretarsus IV nearly one third total length of tarsus IV; coxa IV short, sub- quadrate, slightly wider than long; tarsus IV quadrate, ratio of femurogenu IV to tarsus IV approximately 1:1.6. 19. In all mounted specimens of the type series, legs I are twisted so that claw I turns outward rather than inward, resulting in a mesal displacement of certain setae including those of the trochanter. The normal position of legs I therefore probably differs somewhat from closely related forms. THe FamMiIty PYEMOTIDAE 158s Hasirs. Nothing is known of the habits of this mite. Pediculaster ( Acarothorestes) curculionium, new species DESCRIPTION OF NON-GRAVID FEMALE. Length, 229y. (223-290, X = 249.67); width, 96y. (83-110, X = 100.17); distance between anterior sternocoxal condyles of coxae III, 59» (58-79, X = 65.17); punctae covering entire body, including tips of tarsi. Gnathosoma. Small, length 7° (19-26); width 2° (15-17); external dorsals lacking; solenidium 2 small, slightly shorter than solenidium 1, not capitate; distance between first and second palpal setae about 5u. Propodosoma. Dorsum, Distance between internal pseudostig- matal sockets 9y. (9-11, X = 10.66); all propodosomal setae stout, spiculate; stigmatals only three to five y. apart, approximately 25 to 30». long, subequal in size to posterior pseudostigmatals; anterior pseudostigmatals smaller, arising about one areolar diameter pos- terolateral of stigmatals; posterior pseudostigmatals arising nearly directly laterad of pseudostigmata, about 25 to 30y apart. Venter. Five pairs of short, nude, flagellate setae, the two pairs on ventrites II subequal in size, approximately 16 to 21y, distinctly longer than setae of ventrites I; internal ventrals II only slightly closer to ex- ternal ventrals II than to anterior median apodeme; apodemes II straight, strongest laterally, becoming progressively weaker toward the center, incomplete medially, making angle of about 60 degrees with anterior median apodeme. Hysterosoma. Dorsum. All dorsal setae except laterals III, dorsals IV, and all setae of segment V stout, spiculate, their apices blunt or bluntly pointed; dorsals I (21-25, X = 23.60) long, 29p. (28-31, X = 29.67) apart; laterals I about two to four areolar widths anterior to line drawn between dorsals I, 28y, (28-33, X = 29.50) long, distinctly closer to respective dorsals than dorsals to each other; dorsals II 22u. (17-23, X = 21.00) long, 42y. (41-45, X = 42.17) apart; dorsals III 25y, (23-26, X = 24.83) long, as far apart as dorsals II; laterals IIT bristlelike, 7». (5-7, X = 6.50) long, arising about one dorsal areolar width laterad of respective dorsals; dorsals IV similar to laterals III in size and conformation, well behind and only slightly mesad of laterals IV, 26y. (25-31, X = 27.67) apart; laterals IV 254 (13-25, X = 15.83) long; internal and external caudals I subequal in size, bristlelike, about four y» long, external caudals II 20. These measurements of the holotype not possible. 154 Tue UNIVERSITY SCIENCE BULLETIN bristlelike but longer, about 16x. Venter. All setae of posterior ventral plate short, nude, flagellate, approximately 10 to 15y. long; internal presternals nearly the length of coxa III anterior to coxa III. Legs (all measurements in y.). Length. Leg I, 56 (55-63, X = 58.50); leg II, 75 (74-80, X= 75.67); leg III, 74 (74-80, X= 76.50); leg IV, 58 (58-68, X = 61.00). Width. Leg I, 23 (16-24, XK = 21.83). leg I, 14 *(i-15, xk = 13116). les: Sie ies Cools X = 10.17); leg IV, 17 (14-19, X = 16.17). Segment lengths. Tr 15995 (24-28) XK = 9517). tita: I.) 23° (23-26) x = 24983) at ieeehs (18-20, X = 18.67); ta II, 81 (81-35, X = 32.56); ti III, 14 (14-17, XK = 15.00). ta Ill, 30 (28-33, X= 30:33): ex TV ais Gis=22. X= 19.33); tr IV, 14 (14-19, X = 16:50). fege, LV 9N(8-Oexe— 8.67); ta TV, 14 (14-17, K = 15.00). Setation, Weg alenicxe0iieain 8T +. 1 specialized, fege 4T, tita 12T + 9S; leg II: cx IT, tr 8T, fege IT, ti 4T + 1S, ta 6T + 1S; leg III: cx IT, tr 2T, fege 1T, ti 4T + 1S, ta 6T; leg IV: cx IT, tr 2T, fege IT, ti 4T + 1S, ta 6T. Tibiotarsus I with five rodlike and four short, stout, solenidia, the latter all in apical third of the segment; claw I sharply bent, a conspicuous tubercle on its inner margin; thumb stout but short; solenidia of tibiae II and III small, clavate, sub-basal; solenidium of tarsus II also sub-basal, short and stout, approximately 4.5 x 3y,; tarsi If and III each constricted suddenly into elongate pedicel nearly one-third total length of tarsus; solenidium of tibia IV short, clavate, sub-basal; tarsus IV with three greatly elongated apical setae, the external distinctly longer than the two internals. DistRiBUTION. Known only from the type locality. Type MareriaL. Female holotype and five female paratypes (nos. 1 to 5) from Palmira-Valle, Colombia (Let.), May 14, 1946, from B. Losada S., on Copturomimus sp. (Curculionidae). Type Repositories. Holotype and three paratypes (nos. 2, 3, 5) in the U. S. National Museum, Washington, D. C. One paratype (no. 1) in the Snow Entomological Museum, The University of Kan- sas. One paratype (no. 4) in the Zoologisches Institut, Universitat Erlangen, Erlangen, Germany. Subgenus Pediculaster, sensu stricto (Figs. 57-58 ) Pediculaster Vitzthum, 1931, Memoires du Musée Royal D’Histoire Naturelle De Belgique, Hors Serie—Résultats Scientifiques du Voyage aux Indes Orientales Néerlandaises, 3 (5):36, Acarina. Type, Pygmephorus mesem- brinae R. Canestrini, 1880, by original designation. THe Famity PYEMOTIDAE 155 DESCRIPTION OF NON-GRAVID FEMALE. Propodosoma. Dorsum. “Fins” of esophageal “finned” structure fine, often difficult to see. Hysterosoma. Venter. Posterior ventral plate entire, not divided transversely between coxae III; posterior ventral plate with five or six pairs of setae; apodemes III and IV always present. Species thought to belong in Pediculaster, s. str.: americanus (Banks, 1905) (Pygmephorus ) americanus var. socotrensis (Hirst, 1921) (Pygmephorus ) heliocopridis ( Vitzthum, 1931) ( Pediculoides ) manicatus (Berlese, 1904) (Pygmephorus ) mesembrinae (Canestrini, 1880) (Pygmephorus) Pwichmanni (Vitzthum, 1923) ( Pediculoides ) A comparison of Hirst’s (1921) drawing of P. americanus with the type of that species shows numerous differences in setation and the two are quite possibly distinct. Hasirs. Little is known of the habits of the members of this sub- genus. Vitzthum (1941), states that the mated but non-gravid females of P. mesembrinae ride upon adult flies and drop off at the oviposition sites where they attack the developing larvae. The same author states that the same species has been shown to attack the gall midge, Cecidomyia saliciperda Dut. in the same manner. According to the above author, female mites identified as P. mesembrinae have been taken from the following flies in Europe: Mesembrina mystacea L., Musca domestica L., Stomoxys calcitrans (L.), Siphona irritans (L.), Strathiomys (sic!) sp., and various Borboridae. The same author described P. heliocopridis from an Asian Heliocopris. In the United States, Banks described P. ameri- canus from Musca domestica L. Other host records may be ob- tained from the list below. SPECIMENS E,xAMINED. Pediculaster americanus Banks, Lectotype female (here designated) and four syntype females from Washing- ton, D. C., Jan. 30, 1915 (U.S. N. M. No. 6984). In addition, eight unidentified species from many localities in the United States and from Mexico (Baja California) and from Haiti, from the following situations or hosts: Musca domestica L., anthomyiid fly (prob. Fannia), thorax of “Muscidae,” thorax of Ephydridae, thorax of Leptocera (Diptera, Borboridae), from “fly,” from Diatraea (Lepi- doptera) in rearing cage, peach orchard soil, in manure, on com- mercial mushrooms, on leaves of Azalea sp., treehole of Fagus 156 THe UNIVERSITY SCIENCE BULLETIN grandifolia Ehrh., nest of Peromyscus truei (Shufeldt), nest of Neotoma albigula Hartley, ground litter in forest of roble and pines (Baja California ). Pygmephorus Kramer (Figs. 59-60 ) Pygmephorus Kramer, 1877, Archiv fiir Naturgeschichte, 43(1):248-259. Type, Pygmephorus spinosus Kramer, by original designation. This genus was erected to contain a single species taken from a mole. Beginning with Canestrini (1880), but due chiefly to the various works of Berlese and Oudemans, the limits of the genus became increasingly broad. Krezal (1958, 1959b) includes all those forms comprising the present author’s subfamily Pygmephorinae except that he has retained Microdispodides Vitzthum, 1914. In this paper the genus is restricted to those forms defined in the diagnosis below. Members of the genus are most closely related to those of Pedicu- laster and Microdispodides. They differ from the former in the shape and size of leg I, the shape and size of seta c of trochanter I, the conformation of the ventral apodemes, the shape and position of the peritremes, and other characters. They differ from the latter in the character of leg I, in the type of integumental engraving, and in other characters. Species of Pygmephorus are known to occur in Europe, South Africa, and North and South America. Several forms occurring on mammals have been found in Iceland, Greenland, and Point Barrow, Alaska. Only the females are known. Diacnosis. Females separable from those of all other genera in the family by the following combination of characters: three pairs of propodosomal dorsals [Kramer, in his original description and illustrations, indicated spinosus to have but two pairs of propodo- somal dorsals. Although the author has not seen the type of spinosus, an inspection of many specimens believed to be that species (or relatives not separable from it by the original descrip- tion) showed all to possess three pairs of propodosomal dorsals. The author therefore concludes that Kramer was either unable to observe the third pair or that spinosus was described from abnormal specimens. It should be further pointed out that because of its wide distribution and host range, it is quite possible that spinosus, THE FAMILY PYEMOTIDAE Sz as now understood, will prove to be several closely related species]; leg I four-segmented; seta c of trochanter I short, semimembranous, bladelike, the apical portion proflexed; claw I large, sessile. DESCRIPTION OF NON-GRAVID FEMALE. Length, 110 to 520y,; width, 75 to 220u; width between anterior sternocoxal condyles of coxae III 55 to 170u; hysterosoma oval to elliptical, its posterior margin broadly rounded or truncate; integument often with large, deep punctae, appearing “pebbled.” Gnathosoma. District, squarish or slightly longer than wide, normally directed anteroventrally or ventrally; palps free, distinct, with one or more apical teeth; chelicerae tiny, indistinct; large, ventral esophageal-pharyngeal pump absent, esophagus usually plainly visible as a simple tube. Propodosoma. Dorsum. Inversely T-shaped, often greatly elon- gated behind pseudostigmata, anterior portion prolonged hoodlike over gnathosoma; margins of circumgnathosomal foramen heavily sclerotized, at least laterally, to form “collar” which encloses basal third to quarter of gnathosoma; posterior margin well covered by anterior margin of first hysterosomal tergum; peritremes widely separated, dorsal and circular, ovoid, or elongate troughlike and terminating anteriorly in propodosomal “shoulders”; esophageal “finned” structures distinct or not, if so then two in number, the anteriormost globose to subquadrate. Venter. Anterior ventral plate with four to six pairs of setae, if ventrites II with two pairs then internal ventrals about as close to anterior median apodeme as to external ventrals; ventral externals I forked or not; coxa I articulated at nearly a right angle to the longitudinal axis of body or distinctly obliquely; apodemes II strong, straight or weakly arcuate, complete medially; secondary transverse suture absent; anterior median apodeme very strong, complete; posterior marginal apodemes strong, complete; pleated conjunctival band indistinct between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment entire, its anterior margin overlapping posterior margin to posterior fourth of propodosoma; fifth segment well developed but usually completely enclosed by fourth in dorsal aspect, bearing three pairs of ventral or apicoventral setae; vulva apical or nearly so, vagina enclosed apically by a dis- tinct, sclerotized, elliptical or globose enlargement, its shape de- pendent partly upon mount; posterior genital sclerite often globose, 7—1367 158 THE UNIVERSITY SCIENCE BULLETIN never triangular and flaplike; anterior genital sclerite distinctly pos- terior to line drawn between posterior margins of coxal foramina IV; median genital sclerite absent. Venter. Posterior ventral plate entire, its hind margin also entire; posterior ventral plate bearing six pairs of setae; external presternal well posterior to and scarcely laterad of internal presternal, arising from or not more than two areolar widths in front of apodemes IV; external poststernal present, varying in position; apodemes III weak, often intermittent, not bend- ing posteriorly to join anterior sternocoxal condyles III; apodemes IV distinct only medially, extending about half way to coxae III, directed toward center of coxal foramen; apodemes V present or absent; posterior median apodeme distinct, arising from vicinity of internal presternals and extending posteriorly at least as far as sec- ond axillaries; band of opisthosomal conjunctival pleats present or absent; opisthosoma distinctly longer than distance between anterior margin of coxa III and posterior margin of coxa IV. Legs. All legs well developed, bearing claws; leg I shorter and thicker than leg II; tibiotarsus I elliptical, quadrate, or subpentag- onal, often enormously enlarged, always distinctly wider than other segments of leg; pinnacula present or not, if present then arising from median portion of ventral surface of segment; large, rounded, mediolateral sclerotized plate present on tibiotarsus I; claw I enormously developed, furnished with an equally well de- veloped bladelike thumb which may have a single tooth along its inner margin; trochanter I with three or (usually) four setae, seta c short, bladelike, semimembranous, the apical portion often pro- flexed; solenidia of tibiae IT and III present or absent; tarsus II with five to seven setae and a solenidium, tarsus III with five to seven setae; claws II to IV simple or (usually) with rounded basal dila- tions; pulvilli If to [IV well developed; coxa III subtriangular or subcordate; coxa IV short-rectangulate to elongate and slightly bulbose basally; tarsus IV slightly to distinctly longer than tarsus III, tapering abruptly; ratio of femurogenu IV to tarsus IV 1:3.0 to 1:5.0. Species thought to belong in Pygmephorus: ceratophyi Krezal, 1959 crassipes Krezal, 1959 dominguezi Athias-Henriot, 1961 erlangensis Krezal, 1959 forcipatus Willmann, 1952 ignotus Krezal, 1959 islandicus Sellinck, 1940 THe Famity PYEMOTIDAE 159 karafiati (Krezal, 1959) ( Microdispodides ) microti Krezal, 1959 soricis Krezal, 1959 stammeri Krezal, 1959 ursulae Krezal, 1959 Hasirs. Little is known concerning the habits of the mites of this group. P. spinosus and its relatives are found in the nests and food caches and on the bodies of a variety of widely distributed small rodents and insectivores. The remaining forms are found in association with Coleoptera (Scarabaeidae, Buprestidae) or with trees. The possibility exists that the latter may more properly be associated with wood-inhabiting insects, at least during the phoretic period. Since the P. spinosus complex differs markedly from the other species of the genus in both structure and habitat, later re- search may prove it to deserve separate status. Females of the spinosus group become physogastric in the same manner as those of Siteroptes, the hysterosoma becoming ovoid rather than globose, however. The swelling may be less pronounced, the largest female seen by this author measuring approximately 700y. long, 400y. wide, and 550y. high. SPECIMENS EXAMINED. At least five species from many parts of the eastern United States and Alaska, and from Canada, Germany, and Surinam, from the following situations or hosts: “mole,” Sorex araneus L., S. cinereus Kerr, Blarina brevicauda (Say), B. b. caro- linensis (Bachman), Cryptotis sp., Microtus ochrogaster (Wagner ), Peromyscus gossypinus (LeConte), bumblebee nest in old lemming burrow, humus under log, under bark of dead tree, from cativo logs (Surinam ), and from dung beetle. Microdispodides Vitzthum (Figs. 61-63) Microdispodides Vitzthum, 1914, Zoologischer Anzeiger 44(7):315-328. Type, Microdispodides wichmanni Vitzthum, 1914, by original designation (= Pediculoides amaniensis Oudemans, 1912). Previously known only from the type species, a second species of the genus is described herein. The genus stands closest mor- phologically to Pygmephorus, which it resembles in possessing three pairs of propodosomal dorsals, in having the peritremes well separated, in that apodemes II are straight or nearly so, because the vagina is encompassed apically within a sclerotized structure which is oval or elliptical in shape, and other characters. Its mem- bers may be separated from those of Pygmephorus by the shape of leg I and by the lack of claws on legs I and IV. 160 THe UNIVERSITY SCIENCE BULLETIN Members of the genus are known only from Africa [Cameroon and Tanganyika (German East Africa)] and Central America (Guatemala). They are phoretic upon beetles. Only the non- gravid females are known. Diagnosis. The non-gravid females are distinguished from those of all other genera in the family by the following combination of characters: three pairs of propodosomal dorsals and six pairs of propodosomal ventrals; seta c membranous and flaplike, proflexed apically; integument, at least in part, with scalelike engravings; leg IV clawless, pretarsus IV nearly as long as tarsus IV. DESCRIPTION OF NoN-GRAVID FEMALE. Length, 135 to 235y; width, 70 to 135y; width between anterior condyles of coxa III approximately 90y.; hysterosoma oval, its posterior margin broadly rounded or truncate; body well sclerotized, portions of dorsal in- tegument appearing scaled. Gnathosoma. Squarish, normally directed anteroventrally or ventrally; chelicerae tiny, indistinct; ventral esophageal-pharyngeal pump absent; esophagus plainly visible as a simple tube. Propodosoma, Dorsum. Irregularly trapezoidal, often elongate posteriorly behind pseudostigmata; anterior portion extending hood- like over gnathosoma; circumgnathosomal foramen anteroventrad, its lateral margins only moderately sclerotized; posterior margin of propodosoma overlapping or weakly overlapped by first hyster- osomal segment; three pairs of propodosomal setae; peritremes (“Augen” of Oudemans, 1915) widely separated, circular or gut- tate, opening anterodorsally, arising near anterolateral angles of propodosoma; esophageal “finned” structures paired, subequal, sub- globose-rhomboidal. Venter. Anterior ventral plate with six pairs of setae; external ventrals I not furcate; coxa I articulated distinctly obliquely to longitudinal axis of body; apodemes IT usually strong, complete, straight or barely sinuate; secondary transverse suture ab- sent; anterior median apodeme distinct, complete or broken briefly in front of apodemes II; posterior marginal apodemes strong and complete; pleated conjunctival band between propodosoma and hysterosoma not visible. Hysterosoma. Dorsum. First segment either overlapping or being overlapped by propodosoma, bearing two pairs of setae; fifth seg- ment well developed but usually enclosed by fourth in dorsal aspect, bearing three pairs of apicoventral setae; vulva apical or nearly so; vagina surrounded by apical, sclerotized, elliptical, pump- THe FAMILY PYEMOTIDAE 161 like enlargement; three genital sclerotizations, the posterior semi- circular in part, the anteriormost the largest, possessing a pair of lateral projections; anterior genital sclerite well posterior to line drawn between posterior margins of coxae IV. Venter. Posterior ventral plate entire, bearing six pairs of setae; hind margin of posterior ventral plate entire, broadly emarginate; external pre- sternal well posterior to internal presternal and mesad or laterad of latter, arising well anterior to apodeme IV; external poststernals present, long, only slightly behind line drawn between posteromesal articulations of coxae IV; apodemes III indistinct or fragmentary; apodemes IV distinct only medially, extending one-quarter to one- third of distance to coxae III, or with additional fragments of apodemes present between medial portions and coxae III; apodemes IV directed toward centers of coxal foramina; apodemes V lacking; posterior median apodeme distinct, arising from vicinity of internal presternals and extending posteriorly to vicinity of second axillaries; band of opisthosomal conjunctival pleats not distinct if present; opisthosoma distinctly longer than distance between anterior margin of coxa III and posterior margin of coxa IV. Legs. Leg I distinctly shorter and thinner than leg I; tibiotarsus I elongate-elliptical, as wide or narrower than other segments of leg; pinnacula lacking; large, rounded, mediolateral sclerotized area absent on tibiotarsus I; claw I lacking; trochanter I with four setae; seta c bladelike, semimembranous, its apical portion pro- flexed; solenidia of tibiae II and III present; tarsus II with six setae and solenidium, tarsus III with six setae; pretarsi II and II mod- erately elongate, about half length of respective tarsi, pretarsus IV more elongate, slightly longer than tarsus IV; claws II and II present and stout with rounded basal dilations, claws IV lacking; coxa III subcordate or short-triangular; coxa IV_ rectangulate, slightly constricted and bulbose basally; tarsus IV about as long as III (excluding pretarsi), subrectangulate; ratio of femurogenu IV to tarsus IV about 1:1.5. Habits. Nothing is known of the habits of these mites. M. amaniensis is known from the Cameroons and from German.East Africa (Tanganyika) upon Polygraphus con- gonus Wichmann and Platypus dispar respectively, at least the former residing beneath the elytra. Both hosts are members of the Scolytoidea, the former belonging in the Scolytidae, the latter in the Platypodidae. M. pholidotus is known only from a Guatemalan species of Metamasius (Curculionidae-Calendrinae ). 162 THe UNIVERSITY SCIENCE BULLETIN Key to the Females of Microdispodides l(a). Small, 135 to 150y; first axillary adjacent to anteromesal margin of coxal foramen III; Africa...............M. amaniensts (Oudemans) 1(b). Larger, 210 to 240; first axillary two to three times its own length mesad of anteromesal margin of coxal foramen III; Central America M. pholidotus, new species, p. 162 Microdispodides pholidotus, new species Dracnosis. In addition to the key characters listed above, M. pholidotus differs from amaniensis in that the external presternals are mesad of the internal presternals, the external poststernals do not nearly reach the hind margin of the body, and in various other characters. DeEscripTION OF Non-GRAviID FEMALE. Length, 223y (212-239); width, 135y, (133-135); distance between anterior sternocoxal con- dyles III, 85y (83-88); propodosomal dorsum with uniformly dis- tributed, coarse, elongate punctae; discs of first three hysterosomal terga with bands of scalelike integumental engravings, broadest on first tergum; anterior three-quarters of tergum IV with dense, elongate punctae, appearing pilose; tergum V and venter of body nearly uniformly punctate, punctae distinctly smaller and rounder than those of propodosomal dorsum. Gnathosoma. Quadrate, length 7! (25y.); width 7! (23-24y.); gna- thosomal dorsals in transverse row, subequal in size; solenidium II large, barely capitate; solenidium I clavate, about four y long; dis- stance between first and second palpal setae about six wu. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets 47y. (45-47); all propodosomal dorsals stout, nude, short, posterior pseudostigmatals distinctly longer than others, 31ly. (29-31), arising directly behind and close to external pseudostig- matal socket; anterior pseudostigmatals arising at a slightly greater distance directly anterior to above socket, shorter than others; stigmatals distinctly mesad of others, approximately 30y. apart. Venter. Internal ventrals I short, lanceolate, median and external ventrals I nude, flagellate, the three setae on each side arising in a straight line; setae of ventrites II linearly arranged as described for ventrites I, nude, flagellate, the externals longer than the others, subequal in size to externals I; internal and mesal ventrals II slightly mesad of internal and mesal ventrals I, externals II slightly laterad of externals I. 21. This measurement not possible from holotype. THE FamMiILy PYEMOTIDAE 163 Hysterosoma. Dorsum. Dorsal setae of first four segments stout, weakly lanceolate, inconspicuously spiculate, setae of segment V short, bristlelike; dorsals I arising behind middle of segment, 19y. (18-19) long, 39y. (37-42) apart; laterals I arising well anterior to dorsals I, 26y. (26-29), distance anterolaterad to respective dorsals nearly equal to that between dorsals; dorsals II subequal to laterals I in length, 80p (78-83) apart; dorsals III like dorsals IH, 36y. (36- 42) apart; laterals III 21y (21-24) long, about six areolar widths anterolaterad of respective dorsals; all setae of segment IV sub- equal in size, 14y. (14-15), dorsals 25y, (25-29) apart, laterals about the length of a seta anterolaterad; setae of segment V subequal in length, about five », distance between internal caudal and respec- tive external caudal II 15y. (15-18); external caudal I distinctly closer to external caudal II than to internal caudal. Venter. First axillaries short, lanceolate, all remaining setae of posterior ventral plate nude and flagellate, variable in length; internal and external presternals subequal in length, externals distinctly mesad of in- ternals and on transverse line with first axillaries; first axillaries about three times their own length mesad of anteromesal margin of coxal foramina III; second axillaries about 30y. long, internal poststernals about half as long, arising barely mesad of external presternals and well behind second axillaries; external poststernals about 40y. long, arising two areolar diameters posterior or postero- mesad of sternocoxal condyles IV; apodemes III and IV fragmented. Legs (all measurements in ».). Length. Leg I. 50 (47-51); leg Il, 72 (72-74); leg Ill, 77 (75-80); leg IV, 110 (108-113). Width. Leg I, 18 (13-14); leg II, 17 (17-21); leg III, 14 (14-15); leg IV, 11 (11-14). Segment lengths. Tr I, 17 (14-17); tita I, 26 (26); tr Il, 22 (22-24); ta II, 24 (24-25); ti ITI, 15 (14-16); ta III, 25 (24-25); ex IV, 88 (33-36); tr IV, 20 (18-22); fege IV, 11 (11-12); ta IV, 33 (33-34). Setation. Leg I: cx 1T, tr 8T + 1 specialized, fege 27, tita 1IT + 10S; leg II: cx IT, tr 3T, fege 2T, ti 4T + 1S, ta 6T + 1S; leg III: cx IT, tr 2T, fege 2T, ti 4T + 1S, ta 6T; leg IV: xml me tege ile tie 4h (ISP). tary. Tibiotarsus I with six rodlike and four short, stout solenidia, the most basal of the latter the smallest; trochanter I with one long seta which reaches basal third of tarsus; solenidia of tibiae II and III small, clavate, basal, that of tarsus II basal, short, subelliptical; tarsi II and III each with blunt, ventral, apical spur, each possessing two stout, spinose setae in addition; coxa IV distinctly arcuate; 164 THe UNIVERSITY SCIENCE BULLETIN sensory peg of tibia IV indistinct; tarsus IV with three greatly en- larged terminal setae, the apical about five times length of tarsus. DistRIBUTION. Known only from the type locality. Type MartertaL. Female holotype and three female paratypes (nos. 1 to 3) from Guatemala (taken in quarantine at Philadelphia, Pa., Feb. 3, 1936, A. B. Wells) from Metamasius sp. Philadelphia No. 29318, U.S.N.M. No. 36-2773. Type Reposirories. Holotype and two paratypes (nos. 2 and 3) in U.S. National Museum, Washington, D.C. One paratype (no. 1) in the Snow Entomological Museum, The University of Kansas, Lawrence, Kansas. Microdispini, new tribe This tribe includes five genera, four of which are new and de- scribed herein. Three (Myrmecodispus, Glyphidomastax, and Per- peripes ) have been found only in association with doryline ants. DESCRIPTION OF NON-GRAVID FEMALE. Gnathosoma. Usually dis- tinctly longer than wide or, if rarely wider than long, then palps free; palps normal or elongate and fitted for piercing. Propodosoma. Dorsum. With one (usually) or two pairs of setae, if two pairs, then second pair visible only under high (1000X ) magnification; peritremes widely separated, ovate, elongate-ellipti- cal, or troughlike and cylindrical, never circular; three esophageal “finned” structures always present, the center one enlarged, the posterior reduced, never nearly equalling it and usually smaller than or as large as anterior. Venter. Anterior ventral plate always with four pairs of setae; external ventrals I not furcate. Hysterosoma,. Dorsum. First hysterosomal tergum often pro- longed anteriorly to cover most or all of propodosoma, if so, then margins both free and radially striated or not; segment V with two (usually) or three pairs of setae; posterior portion of vaginal wall only rarely sclerotized, never enclosed in enlarged, elliptical to globose pumplike structure; posterior genital sclerite variable, only rarely (in some Microdispus) broadly triangular, flaplike. Venter. Apodemes V absent. Legs. Legs I articulated obliquely to longitudinal axis of body, thinner and usually shorter than (rarely as long as) legs II; claw I often absent, if present, pedicellate; mediolateral “sclerotized plate” lacking on tibiotarsus I; trochanter I with three setae, seta c always setiform; trochanter IV usually shorter than, rarely as long as coxa IV. THe Famity PYEMOTIDAE 165 Key to the Genera of the Tribe Microdispini (Females ) 1(a). Leg IV without claws; tarsi II and III each with a pair of large, basal, subrectangulate setae; palps elongated and acuminate apically; associ- ated with doryline Ante ee ee 2 1(6). Leg IV with claws; tarsi II and III without setae modified as above; palps rarely elongated and acuminate apically; associated with dory- Miryam ts OTeN O bicoreereres aes aie a orn ace crease aes Saas marten weet 3 2(a). Body elliptical in dorsoventral aspect; leg IV at least 20u longer than leg III; posterior margin of first hysterosomal tergum with about 12 lbroadstectlig iste cece ce ceca tence Glyphidomastax, new genus, p. 184 2(b). Body circular to sublimuloid in dorsoventral aspect; leg IV not more than 15y longer than leg III; posterior margin of first hysterosomal tergumelackimestectiniws = oamecente hs ean: Perperipes, new genus, p. 188 3(a). Gnathosoma distinctly wider than long; chelicerae enlarged and blade- JED eeaAiceceh hea gia 6 cen ssceeasveale meAMRen rea ca Apa Peponocara, new genus, p. 179 3(b). Gnathosoma never distinctly wider than long; chelicerae usually in- distinct, if distinct then not bladelike..........................4. 4 4(a). Apodemes IV reaching the margins of coxal foramina III; opisthosomal sternum often with scalelike engravings; associated with doryline ants Myrmecodispus, new genus, p. 173 4(b). Apodemes IV never reaching more than, usually much less than four- fifths of distance to margins of coxal foramina III; opisthosomal sternum lacking engravings; associated with doryline ants or not Microdispus, p. 165 Microdispus Paoli As conceived herein, this genus consists of at least eleven species placed in two subgenera, Microdispus and Premicrodispus. Al- though “typical” forms of these two groups differ markedly in their appearance, intergrades occur which are difficult to assign satis- factorily. Largely because of the enlargement of the first hysterosomal tergum, Microdispus was originally placed in the family Scutacari- dae by Paoli. In this paper, the genus is considered to be closely related to species of Pseudopygmephorus and therefore contains the least specialized forms of the tribe Microdispini. Members of the genus differ from scutacarids and resemble most pyemotids in that the first hysterosomal segment does not project beyond the gnathosoma and lacks radial strengthening striations along its margins, because legs II and III are well separated, in that the posterior ventral plate is not well delimited laterally in front of coxae III, because the peritremes are usually dorsal and oval or cylindrical, in the length of the legs, the ventral position of coxal seta IV, the shape of the posterior genital sclerite, and in the loca- tion of the presternals, in that the posterior margin of the posterior ventral plate is often trilobate, by the setiform structure of seta c, the position of apodemes IV, and other characters. The genus presents an observable sequence of modifications of the posterior 166 THe UNIVERSITY SCIENCE BULLETIN ventral plate accompanied by an interesting migration of setae. In most Premicrodispus, the first axillaries are situated nearly upon the coxal foramen. In Microdispus, these setae are progressively farther forward, and in some species may be the length of the coxa anterior to the latter. Apodemes IV, bearing the external presternals, undergo a similar but even more pronounced anterior migration and are often so close to apodemes III that the stem of the posterior median apodeme is very short or virtually absent. Species are known from Europe, North and Central America, and Australia. Only the females are known. Diacnosis. Females of this genus are separable from those of Pseudopygmephorus by the characters of the tribe and from those of Myrmecodispus in lacking an engraved hysterosomal sternum and because apodemes IV do not reach coxal foramina III. DESCRIPTION OF FEMALE. Small mites, length, 140 to 230y; width, 50 to 130u; distance between anterior sternocoxal condyles III, 40 to 75u.; hysterosoma elliptical to oval, its posterior margin broadly rounded or subtruncate; body weakly to moderately sclerotized, lacking integumental engravings, but most lateral portions of pos- terior ventral plate usually with 2 to 12 distinct pleats. Gnathosoma. Well developed, slightly to distinctly longer than wide; palps with one or more apical teeth but not sharply acuminate; esophagus slender, appearing straight or variously spiralled or coiled. Propodosoma. Dorsum. Inversely T-shaped or elongate and trapezoidal, slightly to well developed behind pseudostigmatal sockets; anterior portion of dorsum prolonged hoodlike over gna- thosoma, foramen appearing ventral, only weakly sclerotized, not forming distinct collar; posterior margin to entire dorsum covered by first hysterosomal segment; one pair of propodosomal setae or (rarely), if two pairs, then second pair visible only under high (1000 ) magnification; peritremes anterior, dorsal or dorsolateral, widely separated, diverging anteriorly, oval, elliptical, or elongate and troughlike, if the latter then terminating anteriorly in propo- dosomal “shoulders”; two or three subcontiguous esophageal “finned” structures present, the anterior present or not, if present, resembling a “bow-tie,” the middle usually much the longest, rectangulate, the esophagus distinctly thickened within it, the third squarish, subglobose, or triangulate, often indistinct. Venter. Internal ventrals II arising anterior to external ventrals II, usually closer THE Famity PYEMOTIDAE 167 to them than to anterior median apodeme; external ventrals I not furcate; coxa I articulated at nearly a right angle to longitudinal axis of body or (usually) distinctly obliquely; apodemes II com- plete or incomplete medially, together V- or weakly W-shaped; secondary transverse suture present or absent; anterior median apodeme usually strong and complete, rarely incomplete or absent; posterior marginal apodemes well developed, complete; distinct band of conjunctival pleats absent between propodosoma and hysterosoma. Hysterosoma. Dorsum. Four or five visible segments; first seg- ment overlapping posterior quarter to entire propodosoma, but margins not projecting free, not radially striate; fifth segment well developed, completely covered by fourth in dorsal aspect or not, bearing (usually) two or three pairs of apicoventral setae; vulva apical or apicoventral, often indistinct; apical portion of vaginal wall not sclerotized, not enclosed by sclerotized elliptical to globose enlargement; posterior genital sclerite straplike to broadly tri- angular; anterior genital sclerite distinct, acuminate, variable in position. Venter. Posterior ventral plate entire, bearing five or six pairs of setae, its hind margin entire or tripartite, if tripartite then incisions extending anteriorly to posteromesal margins of coxal foramina IV; first axillaries arising nearly in line with or distinctly anterior to anterior margin of coxal foramina III; external pre- sternals arising directly behind internal presternals or well laterad of them but always on apodemes IV; external poststernals present, not or only slightly posterior to line drawn between posterolateral margin of coxal foramina IV; internal poststernals present or (rarely ) absent; apodemes III absent or present, if present straight, com- plete mesally; apodemes IV distinct only mesally, extending one- half to four-fifths of distance to coxal foramina III, directed towards anterior half of foraminal margin or (rarely) directed well anterior to foramina; posterior median apodeme distinct, extending pos- teriorly to vicinity of coxae IV; distinct band of opisthosomal con- junctival pleats absent; opisthosoma distinctly longer than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV. Legs. Leg 1 thinner than, shorter or as long as leg IT; tibiotarsus I elongate, elliptical to cylindrical, as wide as or narrower than femurogenu I; pinnacula present or lacking, if present then terminal and bearing a pair of rodlike solenidia; mediolateral “sclerotized plate” of tibiotarsus I absent; claw I absent; tarsi IT and III each 168 THe UNIVERSITY SCIENCE BULLETIN with five or six tactile setae; solenidia of tibiae II and III (prob- ably )** present; pretarsi short, one-quarter the length of the tarsi or less; claws II to III with or without basal dilations, distinctly larger than claws IV or not; length of pulvilli variable; coxa II triangular, subcordate or broadly crescent-shaped; coxa IV short- rectangulate to rectangulate, evenly dilate basally or with slight to pronounced basal bulblike dilation but lacking subbasal con- striction; trochanter IV not longer, usually distinctly shorter than coxa IV; femurogenu IV with 1] seta; tarsus IV always distinctly longer than, often twice as long as tarsus III; ratio of femurogenu IV to tibiotarsus IV approximately 1:3.3 to 1:5.5. Key to the Subgenera of Microdispus (Females ) l(a) With the following combination of characters: hind margin of posterior ventral plate entire; leg I distinctly shorter than leg II; apodemes IV meeting posterior median apodeme on or in front of line drawn be- tween anterior sternocoxal condyles III........ Microdispus s. str., p. 172 1(b). Hind margin of posterior ventral plate usually tripartite; leg I shorter than or as long as leg II; apodemes IV meeting posterior median apo- deme on or posterior to line drawn between anterior sternocoxal con- Gyles SDT eee ea ae eae ERNE aN at) Premicrodispus, new subgenus, p. 168 Premicrodispus, new subgenus (Figs. 64-66 ) Type, Microdispus (Premicrodispus ) chandleri, new species. Forms of this subgenus present an interesting morphological tran- sition between certain forms of Pseudopygmephorus on the one hand and forms of Microdispus s. str. and Myrmecodispus on the other. Most Premicrodispus resemble most Pseudopygmephorus in general body shape, in the shape and location of apodemes IV, in the position of the presternals, in having a tripartite posterior ven- tral plate, in the shape and position of the pseudostigmata, in pos- sessing an elongate gnathosoma and elongate legs I, in the shape of the propodosoma, and in other characters. They differ from Pseudopygmephorus in the tribal characters given above. As presently conceived, the subgenus contains at least six species, four of which are previously undescribed. These fall into two distinct species groups. The first of these is most closely related to Pseudopygmephorus and includes four species, all found in decay- ing vegetation. Members of this group are widely distributed, one species (undescribed ) known only from Australia, one from Europe and North America (silvestris Jacot, 1936), one reported only from 22. Not visible on some poorly mounted specimens of Microdispus s. str. THe Famity PYEMOTIDAE 169 Europe (angustus Krczal, 1959) and one restricted to North Amer- ica (chandleri, n. sp.). The second group, associated with insects, resembles certain species of Myrmecodispus in the shape of the hind margin of the anterior ventral plate, in the shape and size of claws II and III, and in having seta I of tibia IV enormously en- larged. Included in this group are two species, one phoretic upon the termite Reticulitermes flavipes in Indiana, the other from the doryline Nomamyrmex hartigi in Panama. The discrepancy in hosts may be explained by the fact that the specimens of N. hartigi were raiding a termite colony when captured. Diacnosis. In addition to the key characters mentioned above, separable from most species of Microdispus s. str. as well as from Myrmecodispus by the shape of coxa IV, which is not or only barely bulbose and distinctly wider at base than at apex. DESCRIPTION OF NON-GRAVID FEMALE. Length, 140 to 210; width, 50 to 90u.; width between anterior sternocoxal condyles III, 40 to 75y.. Propodosoma. Dorsum. Well developed behind pseudostigmatal sockets. Venter. Secondary transverse suture absent. Hysterosoma. Dorsum. First segment overlapping only posterior third to quarter of propodosoma; posterior genital sclerite straplike or tonguelike, anterior genital sclerite well posterior to line drawn between posteromesal margins of coxal foramina IV. Venter. Hind margin of posterior ventral plate usually tripartite; first axillaries arising nearly in line with anteriormost margins of coxal foramina III; apodemes IV meeting posterior median apodeme on or behind line drawn between anteriormost margins of coxal foramina III. Legs. Pinnacula of tibiotarsus I lacking; coxa III triangular to short-subcordate; coxa IV more or less evenly dilated toward the base, not or only barely bulbose basally; tarsus IV not usually twice as long as tarsus III, ratio of femurogenu IV to tarsus IV approxi- mately 1:3.8 to 1:4.6. Previously described forms thought to belong in Premicrodispus: angustus (Krezal, 1959) (Pygmephorus ) silvestris (Jacot, 1936) (Pygmephorus ) Hasrrs. See above. SPECIMENS EXAMINED. P. silvestris (Jacot) and four additional unidentified species from Kansas, Indiana, Panama (C. Z.) and from Western Australia, from the following situations or hosts: Eucalyp- tus sp., punky stump, Nomamyrmex hartigi (Westwood) (Formici- dae, Dorylinae), Reticulitermes flavipes (Kollar) (Isoptera). 170 THE UNIVERSITY SCIENCE BULLETIN Microdispus (Premicrodispus ) chandleri,”°> new species Diacnosis. This species is most closely related to silvestris. It may be separated from that species, as well as from angustus, in that laterals III are spaced about two-thirds their own length from their respective dorsals, and by the three pairs of setae on seg- ment V. DescripTION OF Non-GRAVID FEMALE. Length, 156y. (148-162, X = 153.66); Width 7! (79y.); width between anterior sternocoxal condyles III, 42y. (39-43, X = 41.60); hysterosoma elliptical, weakly sclerotized, indistinctly punctate; posterior ventral plate laterad of legs with three to four longitudinal striae. Gnathosoma. Rectangulate, length, 23». (22-24, X = 22.66); width, 12y. (12-13, X = 12.66); internal dorsals submarginal, the externals small, marginal, usually not visible in dorsoventral aspect; solenidia I and II indistinct, probably absent; distance between first and second palpal setae about three v.; esophagus with series of two to four close spirals between buccal cavity and propodosomal “finned” structure. Propodosoma. Dorsum. Thickly T-shaped; distance between in- ternal pseudostigmatal sockets 25y. (22-25, X = 23.60); single pair of pseudostigmatals appearing to arise on or near anteromesal mar- gin of external pseudostigmatal sockets, these stout, as long as pseudostigmata, sparsely spiculate; peritremes elongate-elliptical, about two and one-half times longer than wide, anterior termina- tions lateral, not reaching anterior propodosomal margin. Venter. All ventral setae nude, flagellate, those of ventrites I subequal, about 12u., externals II 17 to 20y, internals II slightly shorter; apodemes Il together U-shaped, complete; posterior marginal apodemes moderately arcuate. Hysterosoma. Dorsum. Posterior margins of segments I to III emarginate, that of third much more broadly than others; all dorsal hysterosomal setae stout, sparsely spiculate; dorsals I arising about three-quarters of their own length posterior to laterals I, 14y. (X = fe 00) long, 3lp (28-30, X = 29.80) apart; laterals I 25y (23-26, = 23.83) long; dorsals II subequal to dorsals I in size, 26y. (25-26, < = 25.50) apart; dorsals III subequal to dorsals I and II in size, 18y. (15-18, X = 16.16) long, 24y. (22-26, X = 24.00) apart; laterals HII 15y. (13-15, X = 13.83) long, about two-thirds their own length 23. This species is named in honor of Dr. Leland Chandler, Department of Entomology, Purdue University, Lafayette, Indiana. 24. This measurement not possible from holotype. THe FamMity PYEMOTIDAE 171 posterolaterad of dorsals II; setae of tergum IV in transverse row, dorsals IV 20. long (19-21, X = 19.50), only 94 (8-9, X = 8.16) apart, laterals [TV 18y (17-19, X = 17.80) long, about twice as far from respective dorsals as dorsals from each other; setae of seg- ment V equidistant from each other, the internal caudals and ex- ternal caudals II subequal in size, approximately 17 to 19, external caudals I bristlelike, about seven to eight y; vulva seemingly apico- ventral, its posterior extremity marked by a crescentic line; posterior genital sclerite short and tablike. Venter. Posterior ventral plate with five pairs of setae (internal poststernals lacking), all nude, flagellate, about 13 to 19y; external presternals only slightly laterad of external poststernals, in longitudinal line with internal pre- sternals, about six to eight » behind line drawn between postero- mesal margins of coxal foramina IV; first axillaries immediately ad- jacent mesally to anterior sternocoxal condyles of coxae III. Legs (all measurements in y.). Length. Leg I, 37 (37-39, X = 38.00); leg II, 46 (41-47, X = 45.33); leg III, 43 (37-43, X = 41.46); leg IV, 62, (58-64, X = 61.80). Width. Leg I, 8 (8-9, X = 8.16); leg II, 11 (10-12, K = 11.66); leg III, 9 (8-9, K = 8.66); leg IV, 10 (9-11, X= 9.83). Segment lengths. Tr I, 14 (14-15, X=14.17); fatale MS) (XC 1S OO): tre be hy (WAS eX 5:50) tale 1S) (lo. eG .00) tie Wl 7 OX —— 7-00) tay WN Vo) (GIS =a X= 14733) ex DV, 20 (18-20) X = 19°33). tr IV, 17 (15-17, X = 15.40); fege 1V,, 6 (Gap xXe— 6:50) hata Vi 20) (20-2259 — 01,00) Setation.. Legere ex OT, tr 3T, fege 4T, tita JOT + 9S; leg II: cx IT, tr 3T, fege 2T, ti 4T + 1S, ta 6T + 1S; leg III: cx IT, tr 2T, fege 2T, ti 4T + 1S, ta 6T; leg IV: cx IT, tr 2T, fege IT, ti 4T + 1S, ta 6T. Tibiotarsus I with five rodlike and four, short, stout solenidia, the most anterior and lateral of the latter not reaching tip of segment; solenidia of tibiae subbasal; solenidium of tarsus II elongate, about eight »; coxa III distinctly triangulate in ventral aspect; coxa IV quadrate, evenly and distinctly dilate toward the base; tarsus IV short, constricted abruptly three-fifths of distance to its apex; claws II to IV simple, subequal in size, pulvilli II to [TV not longer than claws. DistrripuTion. Known from Michigan, Indiana, and Georgia. Tyre MatertAL. Female holotype and five female paratypes from near Hovey Lake State Game Refuge, Posey County, Indiana, June 6, 1958, E. A. Cross, from rotten stump in wet woods (berlese sample). Purdue University No. PU580605-5. 172 Tue UNIVERSITY SCIENCE BULLETIN Type Reposrrories. Holotype and one paratype (no. 3) in the U. S. National Museum. Two paratypes (nos. 1, 4) in the Snow Entomological Museum, The University of Kansas. One paratype (no. 2) in the Blatchley Collections, Purdue University, and one paratype (no. 5) in the Zoologisches Institut, Universitat Erlangen, Erlangen, Germany. Subgenus Microdispus Paoli s. str. (Figs. 67-68 ) Microdispus Paoli, 1911, Redia (Firenze), 7:267. Type, Diversipes (Micro- dispus ) obovatus Paoli, 1911, by original designation. Typical members of this group show several specializations also found in the family Scutacaridae, a fact which led Paoli to place M. obovatus in the genus Diversipes. These specializations include an anterior migration of the external presternals in addition to the characters mentioned in the following diagnosis. The author fol- lows Krezal in the belief that the genus belongs in the Pyemotidae for reasons already given. The subgenus includes at least five species from Europe, North America, and Australia. Diagnosis. In addition to the key characters, the first hystero- somal tergum is usually elongated anteriorly to cover all or nearly all of the propodosoma, and tarsus IV is usually characteristically attenuate and always elongate (figs. 67-68). DESCRIPTION OF NON-GRAVID FEMALE. Length, 130 to 230y.; width, 80 to 130%; width between anterior sternocoxal condyles III, 45 to 65y.; hysterosoma usually oval, rarely elliptical. Gnathosoma. Always distinctly longer than wide. Propodosoma. Dorsum. Inversely T-shaped; one pair of pro- podosomal setae; peritremes elongate-troughlike, terminating an- teriorly in propodosomal “shoulder.” Apodemes II together, if complete, U-shaped. Hysterosoma. Dorsum. First tergum usually elongated anteriorly to cover entire propodosoma; fifth segment always with two pairs of setae; posterior genital sclerite usually narrowly wedge-shaped, rarely broadly wedge-shaped or straplike. Venter. Posterior mar- gin of posterior ventral plate entire, bearing six pairs of setae; first axillaries arising distinctly anterior to coxae III, usually by more than half the length of the foramen; internal poststernals present; apodemes IV joining posterior median apodeme on or (rarely) above line drawn between anterior sternocoxal condyles III. THE Famity PYEMOTIDAE is Legs. Leg I distinctly shorter than leg II; tibiotarsus I elongate, cylindrical, slightly to distinctly narrower than femurogenu I; tarsi II and III with six tactiles; claws II and III with basal dilations, dis- tinctly larger than claws IV; coxa IV broadly to narrowly rec- tangulate, barely to (usually) greatly dilated bulblike basally; trochanter IV distinctly shorter than coxa IV; tarsus IV elongate, usually characteristically evenly attenuate; ratio of femurogenu IV to tarsus IV approximately 1:3.3 to 1:5.5. Species thought to belong to Microdispus, s. str.: obovatus (Paoli, 1911) ( Diversipes, Scutacaridae ) setosus (Evans, 1952) ( Diversipes) Hasits. Several species contain single, large eggs, suggesting that the size of the gravid female is similar to that of species of scutacarids and extreme physogastry is lacking. Habits are other- wise unknown. SPECIMENS EXAMINED. Five species from the eastern part of the United States and from Australia (West Australia), from the fol- lowing situations or hosts: leaf litter, pine litter, dead grass, mold- ing hemlock and deciduous tree leaves, soil in peach orchard, punky stump in woods. Myrmecodispus, new genus (Figs. 69-71) Type, Myrmecodispus dorylinis, new species. This genus is erected to accommodate five undescribed species of microdispines, all of which are found in association with various Central American army ants. One species (no. 2) exhibits an in- teresting variability of gnathosomal dimensions, the gnathosomal length of some specimens being a fifth longer than that of others. Complete intergradation seemingly occurs between the two ex- tremes. Only the non-gravid females are known. Diacnosis. Females may be distinguished from those of the most closely related genus, Microdispus, by the following combination of characters: apodemes IV reaching the anteriormost margins of coxae III or nearly so; distinct subapical pinnacula present on tibiotarsus I. In addition, tarsus IV is very elongate, usually at least twice the length of tarsus III. DeEscrIPpTION OF NON-GRAVID FEMALE. Length, 140 to 235y,; width, 90 to 125.; width between anterior sternocoxal condyles of coxae III, 58 to 95y; hysterosoma elliptical to oval, its posterior margin 174 THE UNIVERSITY SCIENCE BULLETIN broadly to rather sharply rounded; body moderately to distinctly sclerotized, opisthosomal venter usually with scalelike integumental engravings; most lateral portions of posterior ventral plate with 2 to 12 distinct pleats. Gnathosoma. Longer than wide, often elongate and acuminate; palps with or without distinct apical teeth, sharply acuminate or not; chelicerae minute and indistinct to moderately distinct; esophagus plainly visible as a straight, simple tube. Propodosoma. Dorsum. Quadrate-trapezoidal, not usually well developed behind pseudostigmatal sockets; anterior portion of dor- sum prolonged hoodlike over gnathosoma, foramen appearing ven- tral, its lateral margins only weakly sclerotized, not forming dis- tinct collar; posterior margin well covered by first hysterosomal segment; pseudostigmatal sockets arising from posterior third of dorsum; one pair of proposodosomal setae; peritremes widely sep- arated, diverging anteriorly, ovate to ovate-elongate; two or three propodosomal “finned” structures present, the anterior resembling a “bow-tie,” the middle always much the largest, rectangulate, the esophagus distinctly thickened within it, the third small, similar in size to the first. Venter. Ventrals I in nearly straight line across plate; internal ventrals II arising distinctly anterior to external ven- trals II, distinctly closer to them than to anterior median apodeme; external ventrals I not forked; coxa I articulated distinctly ob- liquely to longitudinal axis of body; apodemes II weakly to mod- erately developed, often incomplete, if complete then together W-shaped or strongly arcuate; secondary transverse suture lacking; anterior median apodeme weakly to moderately developed, com- plete or incomplete; posterior marginal apodemes strongly arcuate or nearly forming a straight line, prominent to exceedingly well developed, often the most distinct apodemes upon the ventral sur- face; pleated conjunctival band absent between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment overlapping posterior one- quarter to posterior two-thirds of propodosoma; fifth segment al- ways with two pairs apicoventral setae; vulva indistinct; apical portion of vaginal wall not sclerotized, not surrounded by apical, elliptical to globular pumplike enlargement; posterior genital sclerite straplike or tonguelike, anterior genital sclerite distinct, acuminate, well posterior to line drawn between posteromesal margins of coxae THe FAMILY PYEMOTIDAE ie IV or not. Venter. Hind margin of posterior ventral plate tri- partite, incisions extending as far as foramen of coxa IV; posterior ventral plate with six pairs of setae; first axillary arising well anterior to coxa III; external presternal well posterior to internal but arising distinctly lateral to it, always on apodeme IV; external poststernal present, arising from posteromesal margin of coxal fora- men or nearly so; apodemes III complete medially, weakly to moder- ately developed, always distinctly weaker than apodemes IV; apodemes IV well developed, complete medially and joining an- terior margin to anterior one-fourth of coxal foramina III, usually joining posterior median apodeme at or above line drawn between anterior sternocoxal condyles; posterior median apodeme distinct, joining apodemes III anteriorly and extending posteriorly at least as far as line drawn between second axillaries; distinct band of opisthosomal conjunctival pleats absent; opisthosoma distinctly longer than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV. Legs. Leg I distinctly shorter and thinner than leg II; tibiotarsus I elliptoid, not wider than femurogenu I; one subapical pinnaculum present, bearing two or three rodlike sensory setae; mediolateral sclerotized plate lacking on tibiotarsus I; claw I absent to moder- ately developed, if present then always distinctly pedicellate and without thumb; tarsi II and III each with six tactile setae, solenidia of tibiae II and III present; pretarsi short, one-quarter length of tarsi or less; claws II and III stout and with distinct basal dilation, distinctly stouter than those of IV; pulvilli II to IV extending dis- tinctly beyond claws; coxa III shortly subcordate to thickly crescent- shaped in ventral aspect; coxa IV short-rectangulate to rectangulate, not or barely bulbose basally; tarsus IV always very elongate and thin, nearly straight or tapering evenly, slightly less than to dis- tinctly more than twice as long as tarsus III; ratio of femurogenu IV to tarsus IV, 1:4.3 to 1:9.0. Hasits. Little is known of the habits of these mites. To judge from the samples at hand, three species are apparently restricted in their host relationships to the single species Nomamyrmex esen- becki. The remaining two are found with more than one species of ant, dorylinis usually with Eciton vagans but rarely upon E. hama- tum, and Myrmecodispus species No. 1 from E. vagans and E. dulcius. 176 THE UNIVERSITY SCIENCE BULLETIN SPECIMENS EXAMINED. In addition to the type species as follows: Species No. 1. Panama (Canal Zone): Barro Colorado Island, 399 9 from Eciton dulcius Forel (Rettenmeyer Colony No. E- 204); same, 1¢@ from E. vagans (Olivier) (Colony No. E-154); same, 19 from E. vagans (Colony No. E-240); same, 42 ¢@ from E. vagans (Colony No. E-180); same, 12? 2 from E. vagans (Colony No. E-212). Species No. 2. PanaMa (Canal Zone): Barro Colorado Is., 209 @ from Nomamyrmex esenbecki (F. Smith) (Rettenmeyer Colony No. E-164). Species No. 3. PANAMA: Cerro Galera, 1 @ from Nomamyrmex hartigi (Westwood) (Retten- meyer Colony No. D-118). Species No. 4. PANAMA (Canal Zone): Barro Colorado Is.,2 9 @ from N. esenbecki (Rettenmeyer Colony No. E-164). Myrmecodispus dorylinis, new species DESCRIPTION OF Non-cravip FemMare. Length, 212y, (201-252, K = 215.10 'S. E. 2.43): wadth, 125 (117-125, xX = 12033 ==)saae 0.70); distance between anterior sternocoxal condyles III, 90y. (80- 90, X = 86.42 +S. E. 0.45); body ovate, moderately sclerotized, lacking integumental engravings. Gnathosoma. Trapezoidal (appearing quadrate with palps ex- tended); length, 30% (24-30, X = 27.28+S. E. 0.38); width, 24y. (22-24, X — 23.32 + S. E. 0.14); palps palmate, one dorsal tooth dis- tinctly larger than rest; chelicerae of moderate size. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets, 52y. (47-55, X = 49.67 = S. E. 0.62); propodosomal seta about as long as pseudostigmatal pedicel. Venter. All setae moderate in size, nude, flagellate, 18 to 35y. long; internal ventral I only slightly larger than external ventral I, the former about 28y. long; internal ventral II less than width of one areolus of external ventral II anterior to external ventral II; apodemes II weak but complete, together W-shaped; posterior marginal apodemes heavy, straight, their most lateral corners bending dorsally, then anteriorly, to form large scapuloid plates. Hysterosoma. Dorsum. First dorsal segment covering posterior one-third (posterior margin to posterior third) of propodosoma; setae of first segment in nearly transverse row, dorsals about a third longer than laterals and 77y (69-77, X = 74.64 + S. E. 0.37) apart; dorsals II tiny, 9» (8-12, X = 9.37 +S. E. 0.32) long, 67. (58-68, X = 63.26 =S. E. 0.49) apart; dorsals Ill 57s. (51-58) x — 5460 THe Famity PYEMOTIDAE G7 +S. E. 0.37) apart, 69y. (59-75, X = 66.00 + S. E. 0.71) long; lat- erals III short, 18 (18-19, X = 15.32 +S. E. 0.41) long, arising about half width of dorsal areolus laterad of dorsals; dorsals IV very elongate, 97y. (88-99, X = 93.72 + S. E. 0.71) long, 32y. (28-32, K = 30.63 + S. E. 0.18) apart; laterals [V subequal in size to laterals III but laterad of dorsals IV by the width of the dorsal areolus or more; internal caudals about 18y., external caudals II about half the size of the internals and placed about length of dorsals II laterad of latter. Venter. All setae of posterior ventral plate flagellate, 14 to 29y. long, first and second axillaries and external poststernals sub- equal in size, distinctly larger than other three pairs; apodemes III moderately developed, complete medially and extending laterally beyond anteriormost margin of coxae foramina III, there recurving slightly; apodemes IV stout, becoming more so at sides, extending laterally to meet coxal foramina II along anterior margins, meeting posterior median apodeme behind line drawn between anterior mar- gins of coxal foramina; external presternals arising distinctly laterad of internal presternals and larger than them; internal poststernals spaced subequally between second axillaries; external poststernals extending less than half way to posteriormost margin of opistho- soma; opisthosomal venter without scalelike engravings, its posterior margin entire. Legs (all measurements in»). Length. Leg I, 44 (39-44, XK = 49.2458. E. 0.28); leg I, 66 (59-66, K = 62.24 +S. E. 0.30); leg MU 74 (68-745, X = 71.27 = S..0:44). leg IV, 123 (17-124, x — RASS be 0.38), Width. eo ll: (0-5. x == 113-68 2-5. 0.24); leg II, 23 (19-24, K = 22.16 +S. E. 0.23); leg III, 20 (19-22, Ne 2024 2S, i. O18): leg TV 19 GI4=20 Xx — 1792 = S.-W. 0132): Segment lengths. Tr I, 13 (13-14, X =13.48=+S. E. 0.10); tita I, 22) (20-22, X= 2104 == 5. EK. 0:14): ta fh, 19 (519. X= 18.24 sy 0223) Ml le le Xe 28-= Sy Ei 0109) sta lel 21 GIO Xe 8:88) = Sak. Os9) vex lVin36) (30-30 oX —= COG asa ke 0.16); tr IV, 24 (22-24, K — 22.64+S. E. 0.15); fege IV, 10 (9-10, K = 9.64 = S. E. 0.14); ta IV, 48 (47-50, X = 47.68 +S. E> 0.18). Setation. Leg I: cx IT, tr 3T, fege 4T, tita 11S + 10T; leg II: cx IT, tr 3T,-fege 37, ti 4T + 1S, ta GT + IS; leg II]: cx IT, tr 2T, fege 2T, ti4T + 1S, ta 6T; leg IV: cx IT, tr 2T, fege IT, ti 4T + 1S, ta 6T. Legs II and III stout, barely tapering from trochanter to tibia; tarsi II and III short and constricted abruptly; tibiotarsus I ellip- 178 Tue UNIVERSITY SCIENCE BULLETIN toid-tapering, with moderately sized claw borne upon distinct ped- icel; pinnaculum short and broad, bearing two setae; dorsolateral margin of tibiotarsus I with three prominent bullet-shaped solenidia subequal in size and distance from one another, another solenidium present distally and basally to these; solenidium of tarsus II short and broadly clavate; coxa IV not basally bulbose; solenidium of tibia IV broader than that of III; tarsus IV moderately elongate, ratio of length of femurogenu IV to tarsus IV about 1:4.9. VARIATION. Only a comparatively slight amount of variation was noted in this species, perhaps because many of the specimens ex- amined were from the same ant colony. Asymmetrical variation between members of the same pair of caudal setae was found and, in such cases, the average length of the two setae was given as the length. This average is usually surprisingly similar to the individual measurements made from more “normal” specimens. It is thought that much of the variation in the width of the legs was due to the amount of pressure upon them and it was definitely noticed that the thinnest legs were associated with the highest mounts. Although only a few specimens of the former were available, there were no noticable differences between specimens found upon the two hosts E. burchelli and E. vagans. DistRIBUTION. Known only from the type locality. Type MarertAL. Female holotype and five female paratypes (nos. 5, 6, 7, 9, and 10) from Barro Colorado Island, Canal Zone, Panama, August 7, 1956, M. E. and C. W. Rettenmeyer, from Eciton vagans (Olivier). Ten female paratypes (nos. 1, 2, 3, 4, 11, 12, 13, 14, 20, 24) as preceding but collected July 9, 1956. Three female paratypes (nos. 15, 16, 17) as above but collected May 31, 1956. Three female paratypes (nos. 8, 18, 19) as above but collected April 30, 1956. Three female paratypes (nos. 21, 22, 23) as above but collected August 17, 1956, from Eciton burchelli (Westwood ). Type Repostrories. Holotoype and eight paratypes (nos. 1, 2, 5, 6, 20, 21, 22, 23) in the Snow Entomological Museum, The Uni- versity of Kansas, five paratypes (nos. 3, 7, 16, 18, 19) in the U.S. National Museum, Washington, D.C., two paratypes (nos. 9 and 15) in the Zoologisches Institut, Universitat Erlangen, Erlangen, Germany, two paratypes (nos. 4 and 8) in the British Museum (Natural History), two paratypes (nos. 10 and 11) in the Chicago Natural History Museum, two paratypes (nos. 12 and 13) in the THE FAMILY PYEMOTIDAE 179 American Museum of Natural History, New York City, and two para- types (nos. 14 and 17) in the private collection of Carl W. Retten- meyer, Kansas State University. MaterRIAL EXAMINED. Other than the type series, as follows: PANAMA (C. Z.): Barro Colorado Is., 163 ¢ 9° from Eciton vagans as follows: Rettenmeyer Colony no. E-180, 11 @ ¢?; no. E-219, 4 9 9@;no. E-154, 32 9 ?;no. E-212,499 9; no. E-240, 67 9 9. Same, from E. burchelli, Colony no. E-253, 8 2 ¢. Peponocara, new genus ( Figs. 78-80 ) Type, Peponocara cathistes, new species. This genus is erected to include a single new species of relatively unspecialized microdispine mite taken from a polyphagine roach, upon which it is probably parasitic. Superficially resembling a long-legged form near Premicrodispus, the species differs greatly from all other microdispines in the structure and shape of the gnathosoma. Gravid and non-gravid females are similar in appear- ance since, in the manner of Microdispus, the female seemingly ma- tures only one egg at a time. Diacnosis. Females are distinguishable from those of all other microdispine pyemotids by the shape and structure of the spherical, flattened gnathosoma. This structure has expanded anteriorly and laterally, with a consequent diminution and ventral displacement of the mouthparts. The chelicerae are large and distinctly blade- like, often extruded. Males and immatures are unknown. DESCRIPTION OF NON-GRAVID FEMALE. Length, 175 to 250y; width, 110 to 150y. (these variations due in part to enlargement of hysterosoma in gravid females ); width between anterior sternocoxal condyles of coxae III, 70 to 75y; hysterosoma ovate, its posterior margin well rounded; posterior margin of first hysterosomal seg- ment lacking broad teeth; body moderately sclerotized; posterior portions of dorsal segments and of posterior ventral plate and often of opisthosomal venter with irregular, impressed lines, these usually increasingly distinct posteriorly; striae present on most lateral por- tions of posterior ventral plate but often difficult to see; punctae usually indistinct. Gnathosoma. Well developed, free, greatly expanded laterally and slightly expanded anteriorly to displace and enclose mouth- parts ventrally; palps distinct but greatly reduced, their apices 180 THE UNIVERSITY SCIENCE BULLETIN with at least one broad tooth; chelicerae enlarged, the stylets dis- tinctly bladelike, often extruded, the guttate cheliceral bases prominent dorsally within the gnathosoma, separated by a heavily sclerotized septum that is bifurcate and recurved posteriorly; ventral esophageal-pharyngeal pump absent but posterior portion of bucca enlarged pumplike; esophagus plainly visible as moderately thick, straight tube possessing Z-like dorsal bend at posterior portion of gnathosoma or anterior part of propodosoma. Propodosoma. Dorsum. Outline trapezoidal, about half as long as wide; lateral margins of gnathosomal foramen weakly developed, not forming a distinct collar; posterior margin well defined by apodeme, barely covered by first hysterosomal segment; one pair propodosomal setae arising directly in front of pseudostigmata; peritremes in anterior third, large, guttate and diverging about 150 degrees, opening near dorsolateral angles; three subcontiguous esophageal “finned” structures present, the anteriormost resembling a “bow-tie,” the middle much the largest, rectangulate, the esoph- agus distinctly thickened within it, the posteriormost tiny, undulate. Venter. Ventral setae I nearly in transverse line; internal ventrals II well anterior to external ventrals IJ and much closer to them than to anterior median apodeme; external sternals I not forked; coxa I arising from ventral plate distinctly obliquely to longitudinal axis of body; apodemes II well developed and complete; secondary transverse suture lacking; anterior median apodeme intermittent behind apodemes II; posterior portion of anterior ventral plate not elongate, distance between apodemes I and II subequal to that between apodemes II and posterior marginal apodeme; posterior marginal apodemes straight, well developed, increasingly so laterally; pleated conjunctival band between propodosoma and hysterosoma absent. Hysterosoma. Dorsum. Anterior margin of first segment barely covering posterior margin of propodosoma; fifth segment well de- veloped but usually completely enclosed by fourth in dorsal aspect, bearing two pairs ventral setae; apical portion of vaginal wall not sclerotized, not surrounded by apical, sclerotized, elliptical to globose pumplike enlargement; vulva apical, moderately distinct; posterior vaginal sclerite distinct, irregularly spherical; posterior ventral plate with six pairs setae, its posterior margin entire; first axillaries about one-third length of coxa III anterior to coxae III; external presternals nearly directly behind internals, arising from apodemes IV; external poststernals present, arising barely mesad THe Famity PYEMOTIDAE 181 of and posterior to posteromesal margins of coxal foramina IV; apodemes III indistinctly present, oblique, complete medially; apodemes IV moderately developed, complete medially but short, reaching less than half the distance to the foramina of coxae III, directed toward anterior sternocoxal condyles, meeting posterior median apodeme on level with line drawn between anteriormost foraminal margins; posterior median apodeme short, often faint, arising anteriorly at Y-shaped juncture with apodemes III, terminat- ing posteriorly near line drawn between second axillaries; posterior ventral plate laterad of legs with series of six to ten nearly parallel striations, these often indistinct depending upon mount; band of opisthosomal conjunctival pleats lacking; opisthosoma distinctly longer than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV; opisthosomal venter lack- ing scalelike engravings except occasional irregular lines near hind margin. Legs. Moderately elongate to (leg IV) elongate, slender; leg I much longer than gnathosoma but distinctly shorter and thinner than leg II; tibiotarsus I elongate-elliptical, not wider than femuro- genu I, lacking distinct pinnacula and mediolateral sclerotized area; claw I absent; tarsi II and III each with six normal tactiles; pre- tarsi short, a quarter the length of the tarsi or less, dilated toward the apex; claws II and III stout but short, without inner basal dila- tion; claws IV present but reduced, less than half size of preceding two pairs; pulvilli extending distinctly beyond claws; coxa III elongate, subcordate in ventral aspect; coxa IV rectangulate, slightly constricted at basal third; trochanter IV only slightly shorter than coxa IV; femurogenu IV elongate, about twice as long as wide, bearing one seta; tarsus IV thin and elongate, ratio of femurogenu IV to tarsus IV about 1:3.5. Hasirs. Little is known of the habits of the single extant species. All specimens seen by the author were part of a single series taken from the polyphagine roach Arenivaga apacha (Saussure). The host roach belongs to a small, rather unique group of deserticolous species known in the United States only from the arid southwest. Because of the enlarged and bladelike chelicerae and because the gnathosoma has been torn from many of the specimens seen by the author, it is inferred that the species is a parasite which attaches rather firmly to the host. This belief is further strengthened in that many of the females are gravid. 182 THe UNIVERSITY SCIENCE BULLETIN In gravid females, extreme physogastry does not occur and the various segments remain telescoped within one another although an increase in hysterosomal size is obvious. All those examined pos- sessed only a single, very large egg, this oval, 125 to 143y long, 98 to 115y. wide. The extrusion of this egg through the apical vulva is seemingly impossible, and oviposition must occur either through some secondary opening or through hysterosomal breakdown. The author believes the former to be the case, positions and aspect of the vaginal sclerotizations and of the tissues between them being quite similar to those of Perperipes except for the muscular common oviduct, which was not observed in Peponocara. The distance be- tween anterior and posterior vaginal sclerotizations is about 80 to 90y. in length, large enough to permit passage of the egg. Males and immatures are unknown. Peponocara cathistes, new species DeEscrRIPTION OF NON-GRAVID FEMALE. Length, 178y, (178-241); width, 117». (117-140); distance between anterior sternocoxal condyles III, 74y. (69-75). Gnathosoma. Length, 24y (283-25); width, 39. (35-39); dorsal setae arising well laterad and in anterior fifth, usually appearing marginal; posteriors well posterolaterad of anteriors. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets, 46y. (41-47); propodosomal seta long, 45 to 50y, nude. Venter. Setae rather stout, length about 17 to 40y, setiform or flagellate, nude; ventrals I of similar shape and size, arising nearly in a straight line; internal ventrals II much smaller than external ventrals and arising distinctly mesad of them; a straight line drawn between externals I and II well laterad of internal ventrals II; apodemes II broadly and flatly U-shaped. Hysterosoma. Dorsum. All dorsal setae nude, elongate-flagellate; setae of first segment in straight line; dorsals I distinctly farther from each other than from respective laterals, and longer than them, length, 6ly. (53-63); dorsals II slightly shorter than dorsals I and arising only 20y. (19-23) apart; dorsals IIT 42y. (41-45) apart, —* (39-50) long, laterals II much shorter, —* (17-24), arising well laterad of and equally posterior to dorsals; all setae of segment IV subequal in size, about 25 to 30y, laterals slightly anterior to dorsals and much farther from them than dorsals from each other; 25. This measurement not obtainable from holotype. THE Famity PYEMOTIDAE 183 setae of segment V apicoventral, internal caudals 15y. (14-18) long, as far from each other as from external caudals II, latter 2ly (18- 23) long; external caudals I 8 (8-12), distinctly closer to internal caudals than to external caudals II. Venter. All setae of posterior ventral plate nude, lanceolate-flagellate, about 17y. (internal prester- nals) to about 41ly (external poststernals) long; first axillaries well mesad of anterior sternocoxal condyles of coxae III; second axillaries arising between mid point and posterior third of coxal foramina III; external presternals twice as wide and about one and one-half times as long as internal presternals; internal poststernals arising slightly anterior to line drawn between posterior sternocoxal condyles IV: external poststernals reaching slightly more than half way to pos- teriormost margin of opisthosoma; posterior margin of opisthosomal venter irregularly sinuate, especially mesally. Legs (all measurements in y.). Length. Leg I, 41 (36-41); leg Ii, — °° (55-58 ); leg III, 64 (58-66); leg IV, 108 (108-112). Width. lege la (Ges) leg: TL 12) (9-12) leo TE (7-9). lee TVs 7. (7-8) Segment lengths. Tr I, 14 (14-15); tita I, 20 (19-20); ta II — 26 (20223) artistes G2) ta I 23420523) exe. 26), (252261): tr IV, 22 (19-23); fege IV, 13 (12-18); ta IV, 41 (40-41). Setation. leg I: (cx 19, t 37, fege 41, tita 1117 4 9S; leg II: cx 17, tr 87, fege 3T, ti 4T + 15S, ta 6T + 1S; leg III: cx 1T, tr 2T, fege IT, ti AN elSeatanoly- les IV: exile str petecenmibat Ae 1S5 taggile Tibiotarsus I with four lateral clavate or bullet-shaped solenidia of which the most basal is distinctly the longest, remaining five sensory setae rodlike; solenidia of tibiae If and III arising from near middle of segment; solenidium of tarsus II large, reaching areolus of tactile seta above it; coxa IV slightly constricted at basal third, enlarged dorsally distad of constriction, much wider than other segments of leg; external setae of trochanter and tibia IV distinctly longer than other setae of leg, that of trochanter reaching to tip of tarsus. VARIATION. No unusual] degree of variation was observed in the small series examined. DIsTRIBUTION. Known only from the type locality. Type MarteriAt. Female holotype and four female paratypes (two gravid) from Presidio, Presidio Co., Texas, September 5, 1949, J. H. Russell, from the blattid Arenivaga apacha (Saussure), U.S.N. M. Lot no. 49-19208. 26. This measurement not possible on holotype specimen. 184 THE UNIVERSITY SCIENCE BULLETIN Tyre Reposrrories. Holotype and three paratypes (nos. 1, 3, 4) in the U.S. National Museum, Washington, D.C. Remaining para- type in the Snow Entomological Museum, The University of Kansas. SPECIMENS EXAMINED. Same data as type series, 5 females. Hasits. See generic description. Glyphidomastax, new genus (Figs. 72-74) Type, Glyphidomastax rettenmeyeri, new species. This genus is erected to accommodate a single, rare, previously undescribed species found phoretic upon the two species of North American army ants Neivamyrmex opacithorax (Emery) and N. nigrescens (Cresson). Because of its heavy sclerotization, the shape of tarsus IV, and the character and size of the first hysterosomal tergum, the species superficially appears to belong to the family Scutacaridae. There seems little doubt, however, that it is most closely related to species of Myrmecodispus, also associated with doryline ants. It is, however, much more specialized than mem- bers of that genus, particularly as regards the gnathosoma, the shape of the body, and modifications of the tarsi and tarsal setae. The mite is immediately separable from any scutacarid species in that seta c of trochanter I is setiform, by the shape of the propo- dosoma, in possessing only one pair of propodosomal setae, in that external caudals I are vestigial, because of the shape of apodemes IV and of the anterior ventral plate, in that the posterior genital sclerite is not broadly wedge-shaped, in the shape of the gnatho- soma and palps, in that the posterior ventral plate is not defined laterally in front of coxae III, and other characters. Only the non-gravid female is known. Diacnosis. Females are distinguished from those of all other genera except Perperipes in possessing a pair of enormous, rec- tangulate setae basally on tarsi II and III, by the shape of the gnathosoma, which is long and acuminate, and more especially by the shape of the palps, which are elongate and sharp, as if fitted for piercing. The body is elliptical and heavily sclerotized, unlike that of Perperipes. DEscrRIPTION OF NON-GRAVID FEMALE. Length, 290 to 315y.; width, 160 to 185y; width between anterior sternocoxal condyles III, 100 to 108»; hysterosoma elliptical, posterior margin evenly rounded; posterior margin of first hysterosomal segment with comb of about THE Famity PYEMOTIDAE 185 12 broad, short teeth; body well sclerotized, without “pebblings” or scalelike engravings but finely and evenly punctate; most lateral portions of posterior ventral plate without pleats but finely striate. Gnathosoma. Well developed, acuminate, much longer than wide; palps unique, elongate, comprising about two-thirds length of gnathosoma, each terminating in two elongate bladelike teeth; chelicerae indistinct, perhaps degenerate; ventral esophageal-pharyn- geal pump absent; bucca well defined; esophagus plainly visible as a simple, thick, straight tube, not spiralled or coiled within a mem- branous pouch. Propodosoma. Dorsum. Anterior portion rounded, sides diverg- ing posteriorly, posterior portion of plate indistinct; anterior portion prolonged hoodlike over gnathosoma, circumgnathosomal foramen ventral, its lateral margins only weakly sclerotized, not forming dis- tinct collar; first hysterosomal segment covering entire dorsum; one pair of propodosomal setae; peritremes large, in anterior half, widely separated, elongate-oval, opening laterally; three subcontiguous esophageal “finned” structures present, the middle one greatly en- larged, its fins uneven and crooked, the anterior and posterior ones reduced. Venter. Ventrals I not in a straight transverse line; in- ternal ventrals II arising well anterior to external ventrals and much closer to them than to anterior median apodeme; external ventrals I not forked; coxa I articulated to anterior ventral plate at angle oblique to longitudinal axis of body; apodemes II weakly to mod- erately developed, each present laterally and medially but absent between; secondary transverse suture lacking; anterior median apo- deme moderately developed, broken two or three times posterior to junction of apodemes II; posterior marginal apodemes greatly en- larged, ten times width of anterior median apodeme, together form- ing an arc; pleated conjunctival band absent between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment semicircular and expanded anteriorly, completely covering propodosoma, margins projecting free and with radial striae, covering base of gnathosoma and basal two or three segments of legs I and II; fifth segment bearing three pairs of ventral setae, the middle pair degenerate; vulva apicoven- tral, indistinct; apical portion of vagina not surrounded by sclero- tized, pumplike enlargement; posterior genital sclerite well internal, irregularly shaped; anterior genital sclerite acuminate, its anterior margin on line drawn between posteromesal margins of coxae IV. 186 THe UNIVERSITY SCIENCE BULLETIN Venter. Posterior ventral plate bearing six pairs of setae, greatly elongated anteriorly; hind margin of posterior ventral plate entire; external presternals well posterior to and slightly mesad of internal presternals, arising slightly in front of apodemes IV; external post- sternals present; apodemes III absent or weak and hidden beneath greatly enlarged posterior marginal apodemes of anterior ventral plate; apodemes IV greatly enlarged, straight and perpendicular to longitudinal axis of body, complete medially and joining coxal foramina III at their anteriormost margins; posterior median apo- deme arising between internal presternals and terminating pos- teriorly in region of internal poststernals; posterior ventral plate laterad of legs with numerous striae, apical striations less pro- nounced than basal; distinct band of opisthosomal conjunctival pleats absent; opisthosoma distinctly longer than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV; opisthosomal venter lacking scalelike integumental en- gravings. Legs. All legs comparatively short and stout, leg I extremely short, approximately the length of gnathosoma, distinctly narrower than leg II; tibiotarsus I subtriangulate, not wider than femurogenu I; distinct pinnacula and mediolateral sclerotized plate lacking on tibiotarsus I; claw I absent; tarsi I] and III each with three normal tactiles and two enormous tactiles, the latter heavily sclerotized, rectangulate, bearing apical prickle; solenidia of tibiae II and III present; pretarsi short, one-quarter length of tarsi or less; claws II and IIT stout, with distinct basal dilations, claws TV lacking; pulvilli II and III extending distinctly beyond claws; coxa III subcordate in ventral aspect; coxa IV subrectangulate, slightly narrowed basally; trochanter IV less than half as long as coxa IV; femurogenu IV bandlike, bearing one seta; tarsus IV elongate and thin, evenly at- tenuate or nearly so, much less than twice as long as tarsus III; ratio of femurogenu IV to tarsus IV, 1:10.0. Glyphidomastax rettenmeyeri, new species DESCRIPTION OF NoN-cGRAvID FEMALE. Length, 296y (315); width, 159». (186); distance between anterior sternocoxal condyles III, 101y. (111). Gnathosoma, Length, 50y. (51); width, 19» (20); distance be- tween palpal setae, lly (12); posterior dorsal setae well behind and Jaierad of anterior dorsals (or appearing directly behind). THE FamiLty PYEMOTIDAE 187 Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets, 48». (52); propodosomal seta short and nude, Qy. (8), situated on prominence about its own length anterior to pseu- dostigmatal organ; apodeme delimiting posterior margin of plate lacking. Venter. Setae moderate to stout in size, length, 28 to 55u. flagellate, nude to sparsely plumose; internals I about 55y., half again as large as externals I; internals and externals IT subequal in size, about 43y, both pairs approximate to coxal foramen II, in- ternals nearly as far from externals II as from externals I, anterior ventral plate elongated behind apodemes II, distance between apodemes I and II much less than that between apodemes II and posterior marginal apodemes. Hysterosoma. Dorsum. Setae of first segment slender, flagellate, dorsals weakly plumose, slightly shorter than and posterior to nude laterals; setae of segments IT to V stout and weakly plumose; dorsals II 88». (85) apart, 55 (63) long, located above centers of coxae III; dorsals IIT 70y. (76) apart, 74y. (78) long, laterals III well anterior and lateral of dorsals, nearly marginal, 66y. (66) in length; setae of segment IV spaced subequally, dorsals 68, (72) long, laterals longer, 88y. (96); setae of segment V distinctly ventral, in- ternal caudals 52u, (52), external caudals I rudimentary, their areoli contiguous with those of internals; external caudals I] about two areolar widths laterad of internal caudals, 33 (31) in length. Venter. All setae of posterior ventral plate nude, flagellate, 22 to 50u; first axillaries anterior to coxal foramina III by about length of coxa III; apodemes IV forming a straight line; second axillaries arising from posterior third of margin of coxal foramina III; external presternals about two-thirds length of internal presternals; internal poststernals close together, straddling posterior median apodeme, arising anterior to second axillaries; external poststernals extending less than half way to posteriormost margin of opisthosoma; opis- thosomal venter without scalelike engravings, its posterior margin entire. Legs (all measurements in ».). Length. Leg I, 48 (48); leg II, 86 (90); leg III, 88 (90); leg IV, 120 (123). Width. Leg I, 19 (20); leg II, 80 (30); leg III, 22 (24); leg IV, 21 (22). Segment lengths. Tr I, 17 (15); tita I, 25 (24); ta II, 39 (387); tr IV, 20 (19); fege IV, 4 (5); ta IV, 55 (55). Setation. Leg I: cx 1T, t 3T, fege 4T, tita 11T + 9S; leg II: cx 1T, tr 3T, ti 41 +15, ta 5T + 1S; leg III: cx IT, tr 2T, fege 2T, ti 4T + 1S, ta 5T; leg IV: cx IT, tr 2T, fege 1T, ti 4T + 1S, ta 6T. 188 THe UNIVERSITY SCIENCE BULLETIN Tarsi If and III conical and elongate; all but one sensory seta of tibiotarsus I setiform or rodlike, the exception lateral, clavate; solenidium of tarsus II large and bullet-shaped, bearing apical prickle; the two setae of trochanter IV and two of tibia IV sub- equal in width, greatly enlarged, wider than any other setae of leg. Hasits. Nothing is known of the habits of this species. DisTRIBUTION. Known only from the type locality. Type MatertaL. Female holotype and one female paratype from The University of Kansas Natural History Reservation, Lawrence, Kansas, April 10, 1957, C. W. Rettenmeyer, from body of Nei- vamyrmex opacithorax (Emery ). Type Repositories. Holotype and single paratype in Snow Entomological Museum, The University of Kansas. OTHER SPECIMENS EXAMINED. Kansas: Lawrence, 19 from Neivamyrmex nigrescens (Cresson). Perperipes, new genus (Figs. 8-9, 75-77 ) Type, Perperipes ornithocephala, new species. This myrmecophilous genus is erected to accommodate three closely related forms apparently restricted to Central American doryline ants. The three forms concerned are separable only in relatively minor size and setal differences. Upon superficial examination, they resemble scutacarid mites quite closely and are the members of the family Pyemotidate most likely to be misplaced into that family. In reality, Perperipes is most closely related to the microdispine genus Glyphidomastax, resembling that genus in the shape and placement of the peritremes, in the structure of the esophagus, in the conformation of the palpi and reduction of the chelicerae, in the structure of the esophageal “finned” structure, in the modified basal setae of tarsi II and III, and in the enlargement of the first hysterosomal segment. It dif- fers from Glyphidomastax in body shape and in numerous other structural modifications. The genus is typically microdispine in possessing but one pair of dorsal propodosomal setae and two pairs of setae on segment V, in that the gnathosoma is elongate and be- cause seta c of trochanter I is unmodified. It resembles members of the family Scutacaridae in body shape, in having an enlarged first hysterosomal segment which completely covers the propodo- soma and which projects slightly over the base of the gnathosoma Tue Famity PYEMOTIDAE 189 and the basal segments of legs I and II, and whose margins are free anteriorly and laterally, as well as in the modification of legs IV. The genus differs from all or most scutacarids in that seta c of trochanter I is unmodified, in that the seta of coxa IV is basal rather than lateral, because the posterior genital sclerite is not def- initely triangular in outline, in the shape and location of the stig- mata, because the external presternals arise from apodemes IV, in that the propodosomal dorsum bears only one pair of setae, because segment V has only two pairs of setae, because pretarsus IV is not elongate, in lacking pinnacula on tibiotarsus I, in the shape of the propodosoma, and because females become enormously physogastric when gravid. Both gravid and non-gravid females are known. Diacnosis. Non-gravid females separable from those of all other pyemotid genera because the body is circular or sublimuloid in dorsoventral aspect. Further separable from all genera except Glyphidomastax in that tarsi I and III each possess a pair of greatly enlarged, subrectangulate basal setae and because the gnathosoma is greatly elongated, the palps fitted for piercing. Gravid females differ from those of other known genera in that the physogastric hysterosoma becomes elongate-cylindrical, possessing pseudoseg- ments (see p. 61). DEscrIPTION OF NON-GRAVID FEMALE. Length, 200 to 250y,; width, 200 to 215y; width between anterior sternocoxal condyles of coxae III, 90 to 100u; hysterosoma broadly ovate, mites nearly circular or circular-limuloid as seen in dorsoventral aspect; posterior margin of first hysterosomal segment lacking broad teeth; body well sclero- tized, without scalelike or “pebbled” integumental markings, not striate on most lateral portions of posterior ventral plate but heavily and finely punctate. Gnathosoma. Greatly elongate and much longer than wide, re- sembling skull of bird; palps distinct, greatly elongate, comprising about three-fourths length of gnathosoma, bladelike, simple, lack- ing distinct apical teeth; cheliceral bases distinct, converging api- cally, but blades degenerate and not fitted for piercing; bucca well defined; esophagus plainly visible as simple, thick, straight tube, not spiralled or coiled within a membranous pouch. Propodosoma. Dorsum. Outline indistinct and variable depend- ing upon mount but much broader than long, anterior margin more or less evenly rounded, prolonged hoodlike over gnathosoma, fora- 8—1367 190 THe UNIvERSITY SCIENCE BULLETIN men ventral but only weakly sclerotized, not forming distinct col- lar; posterior margin indistinct; entire dorsum covered by first hysterosomal tergum; one pair of proposodsomal setae arising laterad of pseudostigmata; peritremes in anterior third, large, ovoid, widely separated, opening near anterodorsal angles of propodosoma, often appearing to open in propodosomal “shoulders,” depending upon mount; esophageal “finned” structures characteristic, the mid- dle one much the largest, its “fins” uneven and crooked, the anterior one “bow-tie’ shaped, distinctly longer than the posterior one. Venter. Ventrals I not nearly in a transverse line; internal ventrals II well anterior to externa] ventrals If and much closer to them than to anterior median apodeme, both pairs approximate to coxal foramina II; secondary transverse suture lacking; anterior median apodeme moderately developed, complete; posterior portion of anterior ventral plate not elongate, distance between apodemes I and II only slightly less than that between apodemes II and pos- terior marginal apodemes; posterior marginal apodemes well de- veloped, increasingly so laterally, mesal portions about same thick- ness as posterior median apodeme; pleated conjunctival band between propodosoma and hysterosoma indistinctly present dor- sally. Hysterosoma. Dorsum., First segment semicircular, expanded laterally and anteriorly, completely covering propodosomal dorsum and projecting slightly beyond, its free periphery with narrow band of indistinct radial striae; fifth segment bearing two pairs of ventral setae; vulva apicoventral, usually distinct; apical portion of vagina well sclerotized but not surrounded by apical, sclerotized, elliptical to globose pumplike enlargement; posterior genital sclerite well internal, small, irregular or crescent-shaped; two anterior genital sclerites present, the anteriormost broadly omega-shaped, the other irregular. Venter. Posterior ventral plate not elongated anteriorly, bearing six pairs of setae, its hind margin entire; external presternals well posterior to and slightly mesad of internal presternals, arising from apodemes IV; external poststernals present, arising adjacent to posterior sclerotization of coxal foramina IV; apodemes III only weakly present; apodemes IV well developed, complete, arcuate, joining coxae III near anterior sternocoxal condyles, meeting pos- terior median apodeme on level with line drawn between centers of coxal foramina III; posterior median apodeme arising anteriorly between internal presternals, often falsely appearing to be a posterior THe Famity PYEMOTIDAE 191 prolongation of anterior median apodeme; posterior ventral plate laterad of legs without distinct striae or integumental pleats, but thick band of lateral conjunctival pleats present, sometimes indis- tinct; thick bands of opisthosomal conjunctival pleats also present ventrally beneath opisthosomal venter and dorsally between seg- ments III and IV and between IV and V; opisthosoma as long as or slightly shorter than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV; opisthosomal venter lacking scalelike integumental engravings. Legs. Short and stout, leg IV reduced; leg I distinctly shorter than gnathosoma, thinner than leg II; tibiotarsus I subquadrate, not wider than femurogenu I, lacking distinct pinnacula and medio- lateral sclerotized area; claw I medium-sized, borne on thick pedicel, pedicel with distinct thumb; seta e of trochanter I stout and b ade- like, at least twice width of other setae of leg; tarsi II and III each with four normal tactiles and two enormous basal tactiles, the latter sclerotized, rectangulate to clavate, serrate or armed with a prickle apically; solenidia of tibiae II and III present, arising at base of segment; pretarsi short, one-quarter the length of tarsi or less; claws II and III stout, with distinct basal dilation, claws IV lacking; pulvilli If and III extending distinctly beyond claws; coxa III elongate-subcordate in ventral aspect; coxa IV subquadrate; trochanter IV about half the length of coxa IV; femurogenu IV greatly reduced, bandlike, bearing one seta; tarsus IV greatly re- duced, subequal to or shorter than tarsus III, not or only slightly attenuate; ratio of femurogenu IV to tarsus IV, 1:3.5 to 1:5.7. DeEscriPTION OF GrAvip FEMALE. As described for non-gravid female except for the following: physogastric, shape greatly elon- gate, cylindrical or slightly clavate posteriorly, resembling small insect larva; length, 1000 to 1300p, width, 220 to 300y. at widest point; venter to and including ventral insertions of the third ter- gosternal muscles, and dorsum up to and including second dorsal segment bulbose, slightly dilated, lightly sclerotized, resembling head capsule of larval insect, this area sometimes set off by a sclerotized band delimiting its posterior margin; body behind above region with series of seven to nine regularly spaced conjunctival pseudosegments, true dorsal segments three and four located in depressions between these pseudosegments; two to three pseu- dosegments between true dorsal segments II and III, three between III and IV, one to three plus terminal lobe posterior to dorsal seg- 192 THE UNIVERSITY SCIENCE BULLETIN ment IV; segment V located ventrally on terminal lobe; all pseu- dosegments, with occasional exception of terminal lobe, with one or more rows of spicules at least laterally, or spicules indistinct or absent; if present, spicules often of two distinct sizes, coarser most frequent on anteriormost and posteriormost pseudosegments but present ventrally on all pseudosegments, finer most common dorsally and on sides of middle pseudosegments; vagina between anterior and posterior vaginal sclerites well defined, forming a birth pore which appears as a posterior linear prolongation of posterior median apodeme when viewed in ventral aspect; ventral conjunctival pseu- dosternum surrounding vagina (oviporal plate) distinctly convex when viewed in profile. Hasirs. A more lengthy discussion of the habits of these mites will be published by C. W. Rettenmeyer of Kansas State University in a forthcoming paper concerned with the inquilines and parasites of Central American doryline ants. Certain aspects not stressed by that author and certain conclusions drawn from studies of compara- tive morphology and phylogeny of the species will be presented here. A feature of members of the genus shared only with Glyphi- domastax is found in the mouthparts. In these two genera the normally styliform chelicerae are vestigial, their functions almost certainly being assumed by the elongated and _ bladelike palps. Since the latter are movable, they may be spread after penetration of the host tissue to insure a more stable attachment during en- gorgement. The structure and conformation of the gravid female differs markedly in several respects from those described for more primi- tive forms such as Pyemotes and proceeds longitudinally so that the gravid mite resembles a tiny coleopterous or dipterous larva. Gravid females are found only with the larval ants (C. W. Retten- meyer, in litt.). Their food is not known, but it is probable that they are parasites of ant larvae or eggs in the same manner as the females of Acarophenax tribolii and its tenebrionid hosts. It seems less likely that they are merely trespassers competing for food with the host larvae. In either event, the adaptive value of the larval mimicry is obvious. Young (i. e., non-gravid) adult females are phoretic upon worker ants. Males are unknown. THe Famity PYEMOTIDAE 193 In the manner of Siteroptes, eggs in various stages of cleavage are found within the body of the female, some containing well- developed larvae. In no cases, however, were other stages found therein, which suggests that a free larval stage may occur. This inference is strengthened by the fact that the most highly developed larval embryos are found in the region of the muscular common oviduct, through which they probably leave the body. Because of the presence of this structure and of the early embryos surrounding it, a complete hysterosomal breakdown resulting in the mass release of eggs or larvae probably does not occur. Dissection of three specimens indicates that the number of fully developed larvae pres- ent within the female at any one time is small, six to ten, while the total number of mature eggs may reach forty or more. Additional eggs in different, earlier stages of development are also easily seen. If a free-living larval stage exists, this stage must be completed within or near the ant brood during the statary phase of the ants’ cycle or shortly thereafter since the young females ride upon the worker ants. Nymphal stage(s) are thought to be absent, the larva molting directly to the adult as in other specialized Pyemotidae and Scuta- caridae. SPECIMENS EXAMINED. In addition to the type species, as follows: PaNaMA (Canal Zone): Barro Colorado Island, March 1-June 17, 1955, M. E. and C. W. Rettenmeyer, 50 non-gravid females from Eciton burchelli (Westwood) and E. hamatum (Fabricius). Same locality and collectors, August 7, 1956, 8 gravid females from E. burchelli. Same locality and collectors, May 18, 1956, 1 non-gravid female from Nomamyrmex esenbecki crassicornis (Smith). The author has included only specimens from E. hamatum in the descrip- tion of the type species but is aware that later study may show the form from E. burchelli to be conspecific with it. The single female from N. esenbecki is undoubtedly specifically distinct. Perperipes ornithocephala, new species DESCRIPTION OF NON-GRAVID FEMALE. Length, 220y. (207-231, = 220 AGES ee 93) width, 20M Gl93-219. x == 208.2012 SAE, 1.94); distance between anterior sternocoxal condyles III, 96u (96-99, X = 97.27=S.E. 0.42). Gnathosoma. Length, —** (55-59, X =57.60+S.E. 0.68); width, — *" (24-26, X = 24.75 = S. E. 0.03); distance between pal- 27. This measurement not possible on holotype. 194 THe UNIVERSITY SCIENCE BULLETIN pal setae about 20y.; external dorsals tiny, setiform, arising postero- laterad of internal dorsals by areolar diameter of latter. Propodosoma. Dorsum. Distance between internal pseudostig- matal sockets, 85y. (85-102, X = 91.80 + S. E. 1.55); propodosomal seta nude, distinctly shorter than pseudostigmatal pedicel. Venter. Setae moderate in size, 23 to 36uv. long, flagellate, nude or sparsely plumose, internals I only slightly longer and wider than externals I; internal ventrals II well anterior to externals and only slightly mesad of them, straight line drawn between externals I and externals II passing barely laterad of internals II; apodemes II evenly arcuate, continuing laterally around anterior margin of coxal foramina II. Hysterosoma. Dorsum. All dorsal setae sparsely plumose; setae of first segment subequally spaced, dorsals arising slightly anterior to laterals, 63y. (63-70, X = 66.92 + S.E. 0.71) apart, 22y (21-25, X = 22.294 +S.E. 0.20) long, laterals about four times as long; dorsals II only slightly longer than dorsals I, 96y (96-108, X = 100.24 + S. E. 0.72) apart, located approximately above center of coxae ITI; dorsals IIT 66y. (66-80, X = 76.41 + S. E. 0.87) apart, 41y. (39-50, X = 41.76 + S. E. 0.58) long; laterals III in nearly transverse line with dorsals, submarginal, 100y (96-107, X = 102.77 +S. E. 1.22) long; setae of segment IV in transverse row, dorsals 52y, (46-59, X = 48.00 + S. E. 0.48) apart, 116 (111-119, K = 116.18 + S.E. 0.43) long; laterals TV submarginal, 127. (120-130, X = 125.60 = S.E. 0.83) long; setae of segment V distinctly ventral, internal caudal 79y. (73-85, X = 80.183 + S.E. 0.75) long, external caudals I lacking, external caudal II arising posterior to internal caudal and about the width of one setal base laterad to it, 39y (36-41, X = 38.45+S.E. 0.48) long. Venter. All setae of pos- terior ventral plate nude or nearly so, flagellate, about 9 to 33y. long; first axillary well laterad of anterior sternocoxal condyle of coxa III; second axillary located at posterior third of foraminal margin of coxa III; external presternal short, less than half the length of internal presternal; internal poststernal arising about half way between anterior median apodeme and external poststernal and behind inferior termination of posterior median apodeme; ex- ternal poststernal extending slightly less than half way to posterior- most margin of opisthosoma; opisthosomal venter without scalelike engravings, its posterior margin entire. Legs (all measurements in y.). Length. Leg I, 46 (46-47, X = THe FAMILY PYEMOTIDAE 195 46.54 +S.E. 0.15); leg Il, 80 (77-83, X = 79.07 +S. E. 0.50); leg Ill, 87 (77-88, X = 81.64 + 1.02); leg IV, 85 (81-93, K = 85.31 + S.E. 0.70). Width. Leg I, 25 (15-28, X= 24.50+5S.E. 0.78); les ll 370 (34-37, X = 36.00-=1S. 20:26), leev 11536 (35-39) Xe— 35.93 +S.E. 0.29); leg IV, 25 (24-29, X= 26.38+S.E. 0.49). Segment lengths. Tr I, 19 (15-19, X= 16.38); tita I, 20 (20-24, X= 22:19 S. FE. 0.95): ta Ill, 33 (29-383, K = 30:63 = 0.16); ui MI, AG (4a xt 1475 == SB 0.19) tal) 33) (29°35, 2X = Sl aye SPE 0:44). ox IV. 337 (60:36) X= 3319 = So. 042); te Iv, 14 IAT! xq 15.95 == SE 093) tege LV, 33-8) Xx 16:29 21S). 0138 adtasliVe 28)(24:30, X= 26:88)==S; EH: 0159) 7) Setation. Keg I: Cx toi fee! 4m tita: loi = 7S leg tly cx Ii, tr 31. tege ol eee Satta ol Selec siiemcxleluetn 2 teses 2M ti it alss ta 6T; leg IV: cx IT, tr 2T, fege 1T, ti 4T, ta 6T (only 3-5 usually visible). Tibiotarsus I with four lateral clavate or bullet-shaped setae; solenidia of tibiae and tarsi II and III basal, often appearing to arise from preceding segments; solenidium of tarsus II] moderately sized, clavate; anterior, enlarged, basal tactile with one large, median, apical tooth and two, tiny, pricklelike lateral teeth (or only large median tooth present); apex of posterior enlarged basal tactile seta similar, this seta not markedly clavate; coxa [TV not bulbose basally; femurogenu IV bandlike, appearing only wide enough to bear stout seta; other segments of leg IV bearing very stout setae of which the external of the trochanter and the external of the tibia are longest, about 150y; ventromedial seta of tibia IV nearly as broad but much shorter, about 105. DistRIBUTION. Known only from the Republic of Panama. Type MarteriaAL. Female holotype and sixteen female paratypes (none gravid) from Barro Colorado Island, Canal Zone, Panama, June 16, 1956, C. W. and M. E. Rettenmeyer, from Eciton hamatum, - Colony No. E-189. Type Reposrrories. Holotype and twelve paratypes (nos. 3 to 14) in the Snow Entomological Museum, The University of Kan- sas. One paratype (no. 1) in the U. S. National Museum. One paratype, (no. 15) in the Zoologische Institut, Universitat Erlangen, Erlangen, Germany, one paratype (no. 16) in the British Museum (Natural History), London, England, and one paratype (no. 2) in the private collection of C. W. Rettenmeyer, Kansas State Univer- sity, Manhattan, Kansas. 196 THe UNIVERSITY SCIENCE BULLETIN SPECIMENS E;XAMINED. PANAMA: Same data as type series, nine females. Pieer River, Darien Province, February 16, 1948, T. C. Schneirla, from Eciton hamatum in nomadic day 2 (Colony No. 148HD), 24 gravid females. Neopygmephorini, new tribe This tribe includes four genera whose interrelationships are un- clear. Three of these groups seemingly possess no entirely unique characters and are therefore difficult to characterize although group differences are obvious upon careful inspection. The fourth genus, Acinogaster, is easily distinguishable from the others although it is more closely related to (and is probably an “exgroup” of) Parapyg- mephorus than any of the other genera are to each other. Members of the tribe are thought to be among the most highly specialized forms of the Pyemotidae and to be ancestral to the family Scutacaridae. Dracnosis. Females easily separable from those of other Pygme- phorinae in that seta c is always modified bladelike, its apex often proflexed, and in having one or (usually) two but never three pairs of propodosomal dorsals. DESCRIPTION OF NON-GRAVID FEMALE. Gnathosoma. Usually quadrate, occasionally longer than wide; palps palmate to subpal- mate, with one or more distinct apical teeth, never greatly elongate, not fitted for piercing; ventral esophageal-pharyngeal pump absent; chelicerae distinct or not, if so then small, aciculate; esophagus not distinctly coiled between pharynx and esophageal “finned struc- tures.” Propodosoma. Dorsum. With one (rarely) or two pairs of setae, if two pairs, both distinctly visible; peritremes widely separated, usually circular, rarely elongate or guttate; esophageal “finned” structures distinct or not, varying in number. Venter. Anterior ventral plate with four pairs of setae; external ventrals I furcate or not. Hysterosoma. Dorsum. First hysterosomal tergum enlarged an- teriorly to cover entire propodosomal dorsum or (usually) not, if so, then margins neither projecting free nor with radial striations; segment V with two (rarely) or three pairs of setae; posterior por- tion of vaginal wall often sclerotized but lacking globose to elliptical pumplike enlargements; posterior genital sclerite usually triangular THe Famity PYEMOTIDAE 197 and flaplike, often broadly so. Venter. Posterior ventral plate en- tire, its hind margin rarely tripartite; apodemes V absent. Legs. Leg I articulated obliquely (usually) or at a right angle to the longitudinal axis of body, size variable; claw I usually pres- ent, only rarely sessile (in some Sicilipes); mediolateral sclerotized area of tibiotarsus I usually present; trochanter I with three setae, seta c semimembranous, bladelike, its apex usually distinctly pro- flexed; piannacula of tibiotarsus I present or absent. Key to the Genera of the Tribe Neopygmephorini (Females ) l(a). First and second axillaries and internal ventrals I and II greatly en- larged, compressed, variously clavate to acuminate Acinogaster, new genus, p. 214 1(6). First and second axillaries and internal ventrals I and II setiform... .2 2(a). Coxa IV broadly constricted about two-thirds of distance from apex, base rounded, bulbose, or, if (rarely) constriction not distinct, then legs long and thin, trochanter IV at least 2.5 times as long as wide (if less than 2.7 times, then trochanter III at least 2.5 times as long as wide); leg I thinner, ?° usually shorter than leg IT Neopygmephorus, new genus, p. 225 2(b). Coxa IV only rarely constricted and bulbose basally, if so then claw I sessilelom lest atleastias widevasiles II, 2s 40994 ye 3 3(a). With the following combination of characters: leg I at least as wide as leg II; longitudinal distance between internal presternal and external presternal always equal to or greater than transverse distance between external presternal and first axillary Pseudopygmephorus, new genus, p. 221 3(b). Leg I usually distinetly thinner and shorter than leg IT, if as wide then longitudinal distance between internal presternal and external prester- nal distinctly less than transverse distance between external presternal andthirstasdllanyaae os ee ee Parapygmephorus, new genus, p. 197 28. As measured at apex of trochanter. Parapygmephorus, new genus This genus contains a number of morphologically heterogeneous but intergrading forms, the most primitive of which resemble closely and are presumably closely related to Pygmephorus. More spe- cialized forms (Parapygmephorus s. str.) closely resemble Neopyg- mephorus. In this paper, the latter genus is considered to be an “exgroup of Pseudopygmephorus for reasons stated elsewhere, and the many morphological similarities between Parapygmephorus and Neopygmephorus are considered to have been derived independ- ently. The arrangement of apodemes IV and of the external pre- sternals, the shape and size of the stigmata, and the occasional pres- ence in both groups of characters universal to the other are instances of such parallelism. 198 THE UNIVERSITY SCIENCE BULLETIN Ecologically, species of Parapygmephorus tend to differ from those of Neopygmephorus, being usually associated with Hyme- noptera, although one form in the author’s collection is found with termites and several species are recorded from decayed wood or sphagnum. It is possible that the latter were dislodged from insects during the processes of collecting and sorting. The genus is here divided into three subgenera, of which Sicilipes is considered the most primitive and Petalomium the most special- ized. Only the females are known. Diacnosis. Females may be separated from those of Pseudopyg- mephorus because the longitudinal distance between the internal and the external presternal is distinctly less than that between the external presternal and the first axillary. In addition, leg I is shorter and usually thinner than leg Il. If (rarely) it is as wide, then claw I is enlarged, opposed by a stout thumb, and tibiotarsus I has a distinct, circular, dorsal, mediolateral sclerotized area. They are distinguished from those of Neopygmephorus in that coxa IV is not distinctly constricted at its basal third, or if so, then claw I is sessile. DerscriPTION OF NON-GRAVID FEMALE. Length, 130 to 430y; width, 70 to 240u; width between anterior sternocoxal condyles III, 55 to 140; hysterosoma ovate, sometimes broadly so; body very weakly to heavily sclerotized, rarely with scalelike integumental engravings; coarse punctae and striae usually absent on most lateral portions of posterior ventral plate. Gnathosoma. Variously shaped, normally directed ventrally or posteroventrally; solenidium II usually greatly enlarged. Propodosoma. Dorsum. Inversely T-shaped to broadly trape- zoidal; anterior portion of dorsum prolonged hoodlike over gnatho- soma, lateral margins of circumgnathosomal foramen thickened to form collar or not; posterior one-fourth to posterior three-fourths covered by first hysterosomal tergum; pseudostigmata located in posterior half; two pairs of propodosomal setae; peritremes large and circular or smaller, transversely ovoid, their mesal margins usually broken or emarginate because of stigmatal openings; peri- tremes located in anterior two-thirds of propodosoma or, rarely, in propodosomal “shoulders”; esophageal “finned” structures distinct or not visible, if distinct, then fins erect, close together. Venter. Internal ventral I arising anterior or posterior to external ventral I; internal ventral II not to distinctly anterior to external ventral II THE FAMILY PYEMOTIDAE 199 and much closer to it than to anterior median apodeme; external ventrals I forked or not; coxa I articulating at nearly a right angle to the longitudinal axis of body or (usually) distinctly obliquely; apodemes II variable, absent to strongly developed, usually incom- plete, strongly arcuate or slightly W-shaped if present; secondary transverse apodeme present or absent, if present, then usually dis- tinct only at sides and continued as thickening mesally; anterior median apodeme weakly to strongly developed, complete or not; posterior marginal apodemes straight, gently arcuate, or lyrate, usually strong, especially at the sides; distinct, pleated conjunctival band present or absent between propodosoma and hysterosoma. Hysterosoma. Dorsum. First segment overlapping posterior one-fourth to posterior three-fourths of propodosoma; fifth segment with three pairs of setae, these apical or apicoventral, or occasion- ally one pair well ventral; vulva apical or apicoventral; vagina en- closed apically by sclerotized, globose to elliptical pumplike enlarge- ment or vagina merely sclerotized apically or soft and indistinct; posterior genital sclerite varying, irregular to triangular and flap- like in shape; median genital sclerite present or not. Venter. Posterior ventral plate with six pairs of setae, its hind margin en- tire, straight, broadly emarginate, or rarely trilobate; position of external presternals variable; external poststernals present, always arising from or close behind (at most half basal width of coxa IV, usually less) posteromesal sclerotization of fourth coxal foramina; apodemes III and IV variable in structure and definition; posterior median apodeme variable in extent and definition; band of opistho- somal conjunctival pleats usually absent, rarely well developed; opisthosoma distinctly longer than distance between anterior mar- gin of coxal foramen III and posterior margin of coxal foramen IV; opisthosomal venter lacking integumental engravings. Legs. Legs elongate or not, usually quite stout, ratio of length to width of trochanter IV, 1:1.2 to 1:3.0, rarely over 1:2.5; leg I distinctly shorter and usually thinner than leg I; tibiotarsus I usually elongate, about one and one-half times to twice as long as wide, distinctly wider than femurogenu I or not; pinnacula present or absent; mediolateral sclerotized area present or not, if present varying in shape and location; claw I dilate basally, variable in size, sessile (rarely ) or pedicellate, pedicel usually broad and constricted basally, if present; thumb variable in size; apical portion of seta c of trochanter I proflexed or not; tarsi II and III each with six 200 THe UNIVERSITY SCIENCE BULLETIN tactiles; solenidia of tibiae II and III present; pretarsus IV usually short, rarely a third as long as tarsus IV; claws II and III with large basal dilations, claw IV simple, usually reduced; pulvilli II to IV well developed, at least slightly longer than claws; coxa III elongate- subcordate to subcordate as viewed in inner ventral aspect, distance between anterior and posterior sternocoxal articulations 22 to 42u.; coxa IV _ short-rectangulate to rectangulate, basa] third not to (rarely) distinctly constricted, basal fifth not to (rarely) distinctly bulbose, if distinctly constricted and bulbose then claw I sessile; ratio, length to width of coxa IV, 1:1.3 to 1:2.3, usually between 1:1.8 and 1:2.0; trochanter IV usually shorter than coxa IV; tarsus IV always distinctly longer than tarsus II, usually elongate and rather evenly attenuate, if short, then constricted abruptly basad of pretarsus; ratio of femurogenu IV to tarsus IV 1:2.9 to 1:10.5. Key to the Subgenera of Parapygmephorus (Females ) 1(a). External ventrals I of anterior ventral plate bi- or trifurecate; solenidium II arising from apex of tibia IT; tibiotarsus I usually with distinct pin- naculum which is at least twice as long as wide Petalomium, new subgenus, p. 209 1(6). External ventrals I of anterior ventral plate not fureate; solenidium IT rarely on apex of tibia IT, usually in normal position or (r arely) arising from tibiotarsal conjunctiva; tibiotarsus I lacking pinnaculum which is;at least twice as! lone as wide. 28 252 es oem a ee 2 2(a). Tibiotarsus I always distinctly wider than femurogenu I, with dorsal, circular, mediolateral sclerotized plate; claw I moderate to large, ses- sile or borne on short pedicel which is as wide or wider than long; ex- ternal presternals arising from apodomes IV or nearly so; secondary transverse suture of anterior ventral plate absent Sicilipes, new subgenus, p. 200 2(b). Tibiotarsus I usually not or only slightly wider than femurogenu I, usually lacking dorsal, circular, mediolateral sclerotized plate—if dis- tinctly wider and plate dorsal and circular, then either external pre- sternals arising well anterior to apodemes IV or secondary transverse suture of anterior ventral plate present; claw I small to moderate, borne on pedicel which is as long or longer than wide Parapygmephorus, sensu stricto, p. 205 Sicilipes, new subgenus (Figs. 81-83 ) Type, Parapygmephorus (Sicilipes) halictinis, new species. This subgenus seemingly stands phylogenetically between Pygme- phorus and Parapygmephorus, s. str., and, with certain Pseudopyg- mephorus, is considered to contain the least specialized forms in the tribe Neopygmephorini. Three undescribed species have been seen, two of which are phoretic upon bees. The third is found in woodland leaf litter. Only the non-gravid females are known. THe FamMILy PYEMOTIDAE 201 Diacnosis. Females easily separable from those of Petalomium in lacking pinnacula, because the tarsal solenidium of leg II is basal, and because external ventrals I are not forked. They are separable from Parapygmephorus by the following combination of characters: tibiotarsus | distinctly wider than femurogenu I and possessing a dorsal, circular, mediolateral, sclerotized plate; external presternals arising from or barely in front of apodemes IV; secondary transverse suture of anterior ventral plate absent. DerscriIPTION OF NON-GRAVID FEMALE. Length, 190 to 225u.; width, 115 to 130; width between anterior sternocoxal condyles III, 80 to 100v.. Gnathosoma. Quadrate or distinctly broader than long. Propodosoma. Dorsum. Inversely T-shaped; distance between internal pseudostigmatal sockets distinctly greater than length of propodosomal dorsum; foraminal collar distinct; posterior margin to posterior one fourth covered by first hysterosomal tergum; pseudostigmata located in posterior one fourth; peritremes without conspicuous rim, circular, widely separated, their mesal margins not distinctly emarginate, located in anterior half of dorsum or in propodosomal “shoulders”; if esophageal “finned” structures dis- tinct, then posterior one enlarged, globose to elliptical, more promi- nent than elongate, cylindrical structure which precedes it, fins erect’ and straight, those of posterior structure much finer than those of anterior. Venter. Internal ventrals I arising anteriorly to external ventrals I; internal ventrals II only slightly anterior to external ventrals II, slightly to distinctly closer to them than to anterior median apodeme; external ventrals I not furcate; apodemes II weak or absent, if present, arcuate and usually complete; secondary trans- verse apodeme absent; anterior median apodeme usually strongly developed, complete; posterior marginal apodeme straight or nearly so. Hysterosoma. Dorsum. First segment. overlapping posterior margin to posterior one fourth of propodosomal dorsum; vulva apical; apical wall of vagina often distinctly sclerotized; posterior genital sclerite variously shaped but never triangular and flaplike; median genital sclerite absent. Venter. Hind margin of posterior ventral plate usually broadly emarginate; external presternals al- ways arising from apodemes IV or less than one areolar diameter in front of them; apodemes III distinct or not, if distinct usually re- curved at the sides to join anterior margins of coxal foramina III; 202 THe UNIVERSITY SCIENCE BULLETIN apodemes IV weakly to strongly developed, directed toward or join- ing coxal foramina III either at top quarter or recurving posteriorly to meet margin behind center; apodemes IV joining posterior median apodeme on or behind line drawn between anterior quar- ters of coxal foramina III; portion of posterior median apodeme be- hind apodemes IV about as long as or slightly longer than portion anterior to apodemes IV. Legs. Ratio of length to width of trochanter IV, 1:2.1 to 1:2.4; leg I thinner than or as wide as leg II; tibiotarsus I usually about one and one-half times as long as wide, always much wider than femurogenu I; pinnacula absent; mediolateral sensory area present, enlarged and circular dorsally; claw I moderate to large in size, if pedicellate, pedicel as wide as or wider than long; thumb moder- ately to greatly enlarged; apical portion of seta c of trochanter I sharply proflexed; basal solenidium of tarsus II in normal position; pretarsus IV never nearly a third the length of tarsus IV; coxa IV subcordate when viewed in inner ventral aspect, distance between anterior and posterior sternocoxal articulations 22 to 30y; trochanter IV subequal to coxa IV in length; tarsus IV moderately elongate, stout basally, tapering evenly to pretarsus or becoming suddenly constricted in apical half; ratio of femurogenu IV to tarsus IV, 1 3.0:to leo 0) Hasits. Nothing is known of the habits of these mites. SPECIMENS EXAMINED. In addition to the type species, as follows: Species No, 1. Highlands Hammock State Park, Florida, 1 9 , leaf litter in low woods (berlese sample). Species No. 2. Moctezuma, Chihuahua, Mexico, 8? @ from thorax of @ Agapostemon sp. ( Apoidea). Parapygmephorus (Sicilipes) halictinis, new species DESCRIPTION OF NON-GRAVID FEMALE. Length, 220» (210-220, X = 215.86); width, 13lp (115-131, X = 126.85); distance between anterior sternocoxal condyles III, 85y. (85-88, X = 86.40); body well sclerotized, coarsely punctate, punctae most noticeable on en- tire hysterosomal dorsum and opisthosomal venter, discs of terga I to III with jagged, shingled engravings besides. Gnathosoma. Length, 25y. (22-25, X = 23.40); width, 28y. (23-28, X = 24.17); external dorsals less than half as long as internals, well posterior to and slightly laterad of them; anterior and posterior palpal setae only three to four y. apart; palp with single, greatly enlarged, mesally directed apicoventral tooth; palpal solenidium II greatly enlarged, cylindrical, solenidium I tiny, acuminate. THE Famity PYEMOTIDAE 203 Propodosoma. Dorsum. Inversely T-shaped, anteriormost por- tion distinctly bulbose; peritremes circular, dorsal, located in anterior one quarter; distance between internal pseudostigmatal sockets 48y. (44-52, X = 47.14); anterior pseudostigmatal seta short, nude, 16y. (11-16, X = 14.50), arising immediately anteromesad of external pseudostigmatal socket, posterior pseudostigmatal seta arising pos- teromesad, very long, 63y. (57-63, X = 58.00), sparsely plumose. Venter. All setae flagellate, sparsely plumose, 33 to 46y. long; in- ternal ventral IT longer than external II, only slightly anterior to it; lateral and ventral margins of circumgnathosomal foramen greatly thickened, forming a distinct gnathosomal collar; apodemes II ab- sent or indistinct. Hysterosoma, Dorsum. All dorsal setae stout, long, sparsely plumose; setae of first segment in transverse row or nearly so, dorsals 41y, (41-47, X = 42.50) long, 52u. (52-57, X := 54.29) apart; laterals I 88y. (85-93, X = 87.83) long, much closer to respective dorsals than dorsals to each other; dorsals II separated by one of their own lengths, 47y. (42-51, X = 46.57) long; dorsals III arising approximately directly behind dorsals II, 50y. (48-53, X = 50.71) long, laterals II] about two areolar diameters posterolaterad, more slender than but about as long as dorsals; dorsals TV 60y. (57-62, X = 59.57) long, only 19, (19-22, X = 20.48) apart, laterals IV nearly that same distance posterolaterad, subequal in length to dorsals; setae of segment V thin, flagellate, sparsely plumose, arising distinctly from the venter, internal caudals shortest, 19y. (17-20, X = 18.14) long, arising well anteromesad of external caudals; ex- ternal caudals I 28. (25-31, X = 27.86) long, much closer to in- ternal caudals than to external caudals II and subequal to or slightly longer than the latter. Venter. All setae of posterior ventral plate flagellate, 20. (1st axillaries) to 50y (external poststernals) long; first axillaries barely anterior of line drawn between anteriormost margins of coxal foramina III, nearly in transverse line with internal presternals; external presternals well behind and distinctly mesad of internal presternals; external poststernals arising distinctly ante- riorly to line drawn between posteromesal margins of coxal foramina IV; apodemes III indistinct; apodemes IV indistinct or distinct only laterally, joining coxal foramina behind middle; posterior margin of opisthosomal venter broadly and shallowly emarginate (to broadly rounded ). Legs (all measurements in ».). Length. Leg I, 55 (54-57, X= 59.43 ); leg II, 69 (64-70, X = 68.17); leg III, 74 (70-85, X = 75.17); leg IV, 126 (118-128, X=124.86). Width. Leg I, 19 (15-19, 204 THe UNIVERSITY SCIENCE BULLETIN K =.17.14): leg IL, 19 (14-19, X=17.48)- lee Ay Gaair X = 14.43); leg IV, 15 (13-15, X = 13.86). Segment lengths. Tr 1 21(19-91, -X = 20.14). tita, 1,924 (24-985 X ole) stile (23-24, X = 23.71); ta Il, 25 (22-25, X = 23.67), tr 111, 24) (24298; X= 96.57). ti W113 (13-14, X 2 13:29). (ta 1 263 (25-23), X = 26.20), cx IV, 29 (25-30, X = 28.00). t IV, 264(265305 98.29); fege IV, 12 (9-12, X = 10.29); ta IV, 41 (87-42, X = 40.86). Setation. Leg I: cx 1T, tr 1T + 2 modified, fege 4T, tita 12T + 8S; leg II: cx IT, tr 8T, fege 3T, ti 4T + 1S, ta 6T + 1S; leg III: cx IT, tr 21. fege 29, tw 47 JS: ta 61; leg IV: cx 1, te 2 tewe etal 4T + 1S, ta 6T. Seta c of trochanter I sharply proflexed, seta d longer but blade- like, undulate; solenidium 3 of tibiotarsus I short, acuminate, soleni- dium 4 laterad of it, less than a third as long, clavate; tibial solenidia of legs II and III arising from basal third, small, clavate; tarsal sol- enidium of leg If resembling solenidium 3 of first leg but smaller; coxa IV short-rectangulate, evenly dilate toward the base, ratio of length to width 1:1.3 to 1:1.6; tarsus IV constricted abruptly in apical half; ratio of femurogenu IV to tarsus IV 1:3.4 to 1:4.7; claw IV simple but well developed, not greatly smaller than claw III. Hasirs. Nothing is known of the habits of this species. The non-gravid females are found, often in huge masses, on the anterior face of the first metasomal tergum of the halictine bee Agapostemon virescens (Fabricius). Occasional stragglers may be found on the discs of the metasomal terga, on the posterior face of the propo- deum, in the mesothoracic spiracular openings, etc. It is probable, in consideration of the numbers of mites involved, that these are acquired by the bee while she is yet in the burrow. DistRIBUTION. Known only from Summit County, Chio, and Tippecanoe County, Indiana. Type MaterrAL. Female holotype from Summit County, Ohio, July 3, 1934, L. J. Lipovsky, from Agapostemon virescens (Fab- ricius). Five female paratypes (nos. 1-5) from Lafayette, Tippe- canoe County, Indiana, July 5, 1932, collector unknown, from Agapostemon virescens. Type Reposirories. Holotype and two paratypes (nos. | and 3) in the Snow Entomological Museum, The University of Kansas, one paratype (no. 2) in the Purdue Entomological Collections, Purdue University, one paratype (no. 4) in the U. S. National Museum, Washington, D. C., and one paratype (no. 5) in the Zoologisches Institut, Universitat Erlangen, Erlangen, Germany. THe Famity PYEMOTIDAE 205 SPECIMENS EXAMINED. In addition to the type material as fol- lows: Lafayette, Indiana, August 25, 1939, collector unknown, numerous females from A. virescens. Same, July 25, 1921, and Sep- tember 28, 1921. Subgenus Parapygmephorus, sensu stricto (Figs. 87-89 ) Type, Paragymephorus (Parapygmephorus) natalensis, new species. This subgenus includes a rather heterogeneous group of fozms, the largest division of which is phoretic upon and perhaps parasitic upon certain African and Asian species of the apo'd subfamily Nomiinae. Other species are associated with other halictine bees, with termites, and with ants. Members of the subgenus exhibit fewer specialized characters than do those of the subgenus Petalo- mium and are most closely related to members of the subgenus Sicilipes. Diacnosis. Females separable from those of Sicilines and Petalo- mium by the characters given in the subgeneric key. DESCRIPTION OF NON-GRAVID FEMALE. Length, 130 to 430; width, 70 to 240u,; distance between anterior sternocoxal condyles III, 55 to 125y; body often with coarse punctae; hysterosomal dorsum oc- casionally with scalelike integumental engravings. Gnathosoma. Quadrate to distinctly longer than wide; palpal solenidium If usually clavate; esophagus usually distinct, occasion- ally with small “bow-tie” finned structure. Propodosoma. Dorsum. Inversely T-shaped to nearly rectangu- late, usually poorly developed behind pseudostigmatal sockets; dis- tance between internal pseudostigmatal sockets usually slightly less than, rarely more than length of propodosoma; foramen usually anteroventrad, rarely ventrad, foraminal collar usually distinct later- ally; pseudostigmata located in posterior third; peritremes circular, rarely somewhat ovoid, located in middle third of dorsum; from one to three “finned” structures usually distinct,?® if so then posterior one subglobose, not enlarged or more prominent than one preced- ing, latter elongate, often greatly so. Venter. Internal ventrals I usually arising distinctly anterior to, rarely posterior to external ventrals J; internal ventrals II slightly to distinctly anterior to ex- ternal ventrals II; external ventrals I not furcate; coxa I attached at an oblique angle to longitudinal axis of body; anterior median apodeme usually strongly developed and complete. 29, If three present, one usually in gnathosoma, two in propodosoma. 206 THE UNIVERSITY SCIENCE BULLETIN Hysterosoma. Dorsum. First segment overlapping posterior one fourth to one third of propodosoma; setae of fifth segment apical or apicoventral; vulva usually enclosed in genital sulcus; apical por- tion of vaginal wall usually sclerotized, sometimes thickly so. Ven- ter. Hind margin of posterior ventral plate straight, broadly emargi- nate, undulate, or rounded; external presternals arising on to well anterior of apodemes IV; distance between external presternals al- ways greater than that between internal presternals; external post- sternals arising half the basal width of coxa IV or less behind line drawn between posteromesal margins of coxal foramina IV; apo- demes III indistinct to distinct, curving posteriorly at the sides to join anterior sternocoxal condyles of coxae III or not; apodemes IV usually joining, always extending at least three-quarters of distance to coxal foramina III; apodemes IV directed toward or joining mar- gins of coxal foramina III at varying points, joining posterior median apodeme on or anterior to line drawn between centers of coxal foramina III; portion of posterior median apodeme behind apo- demes IV about four-fifths to one and three-quarters times as long as that anterior to apodemes IV. Legs. Leg I distinctly shorter and thinner than leg II; tibiotarsus I variable in length, usually about twice as long as wide, as wide as or distinctly wider than femurogenu I; pinnacula absent or reduced, height never nearly twice apical width; mediolateral sclerotized plate variously developed, indistinct, irregular and marginal, or becoming enlarged and circular dorsally; claw I moderate to large in size, pedicellate, pedicel usually about as long as wide; thumb weakly to strongly developed; apical portion of seta c of trochanter I usually strongly proflexed; tarsal solenidium of leg II usually aris- ing from base of tarsus but occasionally from tibiotarsal conjunctiva or from tibial apex; pretarsus ITV never nearly a third the length of the tarsus; basal third of coxa IV not or only slightly constricted, not or only weakly butbose basally; ratio of length to width of coxa IV, 1:1.3 to 1:2.1; ratio of femurogenu IV to tarsus IV, 1:2.9 to 1:6.8. Species thought to belong in Parapygmephorus, s. str.: ?chaetosus (Krezal, 1959) (Pygmephorus ) Hasirs. Nothing is known of the habits of the species included herein. To judge by the presence of bands of pleated conjunctiva, extreme physogastry may occur in at least two species associated with bees. Host preferences are similar to those of the other sub- THe FAamMILy PYEMOTIDAE 207 genera, five of the ten species seen by the author being from bees, two from ants, one from termites, one from decayed wood, and one from “under rock.” SPECIMENS EXAMINED. In addition to the type series as follows: nine species from the central United States and from Brazil, Panama, Mexico (San Luis Potosi), Phillippines (Luzon), and Africa, from the following situations or hosts: Neivamyrmex legionis (F. Smith), N. gibbatus Borgmeier, Nomamyrmex esenbecki (Smith) (all Formicidae, Dorylinae), Nomia sp., Nomia_ philip- pinensis Cockerell, Augochlora sp. (all Apoidea, Halictidae), from under rock, in decayed wood, from workers of Reticulitermes fla- vipes (Kollar) (Isoptera). Parapygmephorus (Parapygmephorus) natalensis, new species DEscRIPTION OF NON-GRAVID FEMALE. Length, 270y. (257-282, X = 269); width, 165y. (162-165, X — 164): distance between ante- rior sternocoxal condyles III, 123y (123-126, X = 125); body mod- erately sclerotized, coarsely punctate, punctae most noticeable on propodosomal dorsum and especially on discs of hysterosomal terga I to IV; other integumental markings absent. Gnathosoma. Length, 37p. (X = 37); width, 36. (31-36, X = 34); external dorsals about three-quarters as long as internals, distinctly more laterad of them than posterior to them; postpalpals spiniform, slightly anterior to internal and slightly laterad of external dorsals; anterior and posterior palpal setae about 5y, apart; palp with distinct apical inner tooth; palpal solenidium I clavate, about half as long as II. Propodosomia. Dorsum. Inversely T-shaped, anteriormost portion distinctly bulbose; peritremes large, circular (or nearly so), dorsal, located about half way between anterior and posterior margins of propodosoma, rather indistinct; stigmatal openings more prominent than peritremes, guttate; distance between internal pseudostigmatal sockets 55y. (55-58, X = 56); anterior pseudostigmatal setae nude, short, lly (11-12, X = 11) long, well anteromesad of internal pseu- dostigmatal socket, posterior pseudostigmatals elongate, 62u. (62-69, X = 66), sparsely plumose, arising on line drawn between internal pseudostigmatal sockets (or nearly so). Venter. All setae flagel- late, sparsely plumose; ventrals I subsequal in size, about 85 to 40y. long; internal ventrals II elongate, about 58 to 63y, externals II only about half as long, arising distinctly posterior to internals; lateral and ventral margins of circumgnathosomal foramen greatly 208 THe UNIvERSITY SCIENCE BULLETIN thickened, forming gnathosomal collar; apodemes II distinct, arcu- ate, terminating at point of origin of external ventrals I; secondary transverse apodemes present, meeting apodemes II medially. Hysterosoma. Dorsum. All dorsal setae stout, elongate (some greatly so), sparsely plumose; setae of first segment in transverse row, dorsals 52y. (52-59, X = 56) long, 63y. (63-68, X = 65) apart; laterals I 163y. (160-163, X = 161) long, much closer to respective dorsals than dorsals to each other; dorsals II 6ly. (61-70, X = 67) apart, 56y. (56-63, X = 59) long; dorsals III arising nearly directly behind dorsals II, 68y (63-75, X—=69) apart, 143, (148-157, X = 148) long, laterals III arising about two areolar diameters posterolaterad, only 30y (30-35, X = 32) long; dorsals IV 97yu (79-105, X = 101) long, only 43y (43-48, X = 46) apart, laterals IV barely anterior to and about three areolar diameters laterad of them, 74. (74-83, X = 80) long; setae of segment V much smaller than those of preceding segments, arising distinctly from ventral sur- face; internal caudals and external caudals J arising in transverse line, their areoli contiguous or nearly so; external caudals I] about three areolar widths anteroventrad of internal caudals; length of internal caudals, 42u, (40-42, X = 41), of external caudals I, 3ly. (X = 31), of external caudals II, 17y (17-18, X = 18). Venter. All setae of posterior ventral plate well developed, flagellate, 30 (inter- nal presternals ) to 48. (external presternals ) long; first axillary dis- tinctly behind internal presternals but well anterior to line drawn between anteriormost margins of coxal foramina III; external pre- sternals well behind and distinctly laterad of internal presternals, arising about one areolar diameter anterior to apodemes IV, ex- ternal poststernals half the basal width of coxa IV behind line drawn between posteromesal margins of coxal foramina IV; apodemes III distinct, at right angles to longitudinal axis of body, extending laterad to longitude of first axillaries; apodemes IV dis- tinct, joining mesal margins of coxal foramina III at or slightly behind their centers; distinct band of conjunctival pleats present (or band indistinct) ventrally beneath posterior ventral plate, dorsally between segments III and IV; vulvar sulcus present; pos- terior genital sclerite elongate, triangular and flaplike. Legs (all measurements in y.). Length. Leg I, 91 (91); leg II, 9 (KX=129):; leg III, 144 (142-146 X= 144). lee Iv, 203 01-203, X = 202). Width. leg I, 24 (23-25, xX = 24) leo es 5-30, X = 27); leg III, 24 (24-25, X = 24); leg IV, 16 (16-22, ). Segment lengths. Tr I, 21 (21-38, X = 26); tita I, 47 7): ty Il, 44 (48-44 X= 44). ta Il, 48 (46-48) x= 47). THe Famity PYEMOTIDAE 209 tr IIT, 51 (51-54, K = 52); ti III, 25 (24-25, X = 25); ta III, 52 (50- 55, X = 52); cx IV, 46 (46-50, X = 47); tr IV, 47 (44-49, X = 47); fege IV, 17 (15-17, X = 16); ta IV, 68 (67-68, X = 68). Setation. Leg I: cx 1T, tr 2T + 1 modified, fege 4T, tita 11T + 9S; leg II: ex 1T, tr 3T, fege 3T, ti 4T + 1S, ta 6T + 1S; leg III: cx 1T, tr ie eve Oly tA Ss tay Oleg Ve (cx Mile tr 2. tego lala ti 4T + 1S, ta 6T. Seta c of trochanter I sharply proflexed, seta d also short and bladelike; solenidium 3 of tibiotarsus I about nine y, attenuate, apex rounded; solenidium 4 slightly more than half as along as 3, rod- like; solenidia 1 and 2 not over eight »; tibiotarsus I elongate, slightly wider than femurogenu; tibial solenidia of legs II and III about five »., weakly clavate, tarsal solenidium of leg II about seven uv, peglike; coxa IV rectangulate, ratio of length to width about 1:1.6 to 1:1.8; ratio, length to width of trochanter III about 1:2.1, tarsus IV evenly attenuate apically, 68y. (67-68, X = 68) long; ratio of femurogenu IV to tarsus IV about 1:4.0 to 1:4.5; ratio, length to width of trochanter IV about 1:2.2 to 1:2.9. Hasits. Nothing is known of the habits of this species. The non-gravid females were taken from the body of the African nomiine bee Crinoglossa natalensis Friese. DistRiBuTION. Known only from the type locality. Type MarTertat. Female holotype and two female paratypes (nos. 1 and 2) from Bluff, Natal, Africa, 9-3-14, collector unknown, from Crinoglossa natalensis Friese. Type Reposirories. Holotype and one paratype (no. 1) in the Snow Entomological Museum at The University of Kansas, remain- ing paratype in the U.S. National Museum, Washington, D. C. Petalomium, new subgenus (Figs. 84-86 ) Type, Parapygmephorus (Petalomium) krezali, new species. This subgenus includes at least a dozen species, most of which are associated with ants. Its members are considered to be highly specialized, as indicated by the characters given below. Dracnosis. Females separable from those of the remaining subgenera in that the body is broadly oval, the propodosoma like- wise broad, in that external ventral I is furcate, in possessing one or two median genital sclerites, because tibiotarsus I usually has a well developed pinnaculum, and in the disproportionate elonga- tion of leg IV, particularly of the tarsus and pretarsus. 210 Tue UNIVERSITY SCIENCE BULLETIN DescripTION OF Non-cravip FEMALE. Length, 210 to 330y; width, 145 to 225u; width between anterior sternocoxal condyles of coxae III, 95 to 145y; body weakly sclerotized, lacking integumental engravings and usually without coarse punctae. Gnathosoma. Quadrate or but little longer than wide. Propodosoma. Dorsum. Broad, distance between internal pseu- dostigmatal sockets at least slightly, usually much greater than length of dorsum; peritremes with mesal margins emarginate, located in middle third of dorsum; esophageal “finned” structures usually in- distinct, single and small to medium in size if distinct. Venter. Internal ventrals II slightly to distinctly anterior to external ventrals II; external ventrals I always distinctly forked, bi- or trifurcate; coxa I articulated at oblique angle to longitudinal axis of body; apodemes II arcuate, weak, terminating laterally at areolus of external ventrals I, usually incomplete medially; secondary trans- verse suture absent; posterior marginal apodemes straight or gently arcuate; distinct, pleated, conjunctival band absent between pro- podosoma and hysterosoma. Hysterosoma. Dorsum, Setae of fifth segment apical or apico- ventral; vulva apical; apical portion of vaginal wall usually scle- rotized and distinct, but enlarged, globose to elliptical sclerotization absent; posterior genital sclerite usually triangular and _flaplike, rarely indistinct; one or (rarely) two median genital sclerites present, irregular or rounded in shape. Venter. External pre- sternals arising on to well anterior of apodemes IV; apodemes III usually indistinct, if distinct, never bending posteriorly at sides; apodemes IV usually extending at least three-quarters of distance to coxal foramina III, rarely less, or indistinct; apodemes IV directed toward or joining coxal foramina III at anterior third or quarter, joining posterior median apodeme on or posterior to line drawn between anterior sternocoxal condyles of coxae III; posterior median apodeme variable, portion behind apodemes IV three-quarters to one and three-quarters times as long as that anterior to apodemes IV; well defined band of opisthosomal conjunctival pleats absent. Legs. Ratio, length to width of trochanter IV, 1:1.2 to 1:2.3; leg I distinctly to greatly shorter than, always thinner than leg II; tibiotarsus I at least twice as long as wide, subequal to distinctly wider than femurogenu I; pinnacula present or vestigial, if present well developed, twice as long as width at apex; medio-lateral sensory area present or absent, if present, then marginal and irregular in shape, never distinctly circular dorsally; claw I moderate in size, THe FamMity PYEMOTIDAE PA pedicellate, pedicel variable in size, thumb absent or moderately developed; apical portion of seta c of trochanter I not or barely proflexed; basal solenidium of tarsus II arising distinctly from apex of tibia; pretarsi IJ and III short, usually one-quarter the length of the tarsus or less; pretarsus IV often distinctly elongate but never more than one-third length of tarsus; coxa III elongate- subcordate when viewed in inner ventral aspect, distance between anterior and posterior sternocoxal articulations 25 to 42y; coxa IV rectangulate, basal third not or only slightly constricted, basal fifth only rarely slightly bulbose, ratio of length to width approximately 1:1.6 to 1:2.3; trochanter IV always distinctly shorter than coxa IV; tarsus IV greatly elongate and thin, usually at least twice length of tarsus III; ratio of femurogenu IV to tarsus IV, 1:5.5 to 1:10.5. Hasits. Nothing is known of the habits of the species included in this genus. One may infer that extreme physogastry does not occur in the females. Seven of the ten forms seen by this author were taken from the bodies of various ants, the other three being found in the soil, in moss, or in woodland litter. Species thought to belong in Petalomium: ?formicarum (Berlese, 1904) (Pygmephorus ) SPECIMENS EXAMINED. In addition to the types species as follows: nine species from the south-central United States and Panama, from the following situations or hosts: Eciton vagans (Olivier), E. mexicanum Roger, E. dulicius Forel, (all Formicidae, Dorylinae), Camponotus sp. (Formicidae, Formicinae), “ants, not Dorylinae,” “under rock, with ants,” frass at base of stump. Parapygmephorus (Petalomium) krezali,?° new species DEscrIPTION OF Non-cGRAvID FEMALE. Length, 296y. (288-362, X = 314); width, 210y. (210-228, X = 221); width between anterior sternocoxal condyles IIT, — *! (162-165); body weakly sclerotized, punctae present but indistinct (to distinct), other integumental markings lacking. Gnathosoma. Length, — *° (31-32); width, — *° (33-34); dorsals stout, sparsely plumose, externals only slightly shorter than internals, arising barely posterior to and well laterad of them; anterior and posterior palpal setae about three py. apart; palpal solenidium II large, capitate, palpal solenidium I tiny, ovoid. 80. This species is named for Dr. Herbert Krezal, Institut fiir Obstkrankheiten, Heidel- berg, Germany, in recognition of his outstanding contributions to the knowledge of this group. 31. This measurement not possible on holotype. 32. This measurement not possible on holotype. 212 THe UNIVERSITY SCIENCE BULLETIN Propodosoma, Dorsum. Broadly inversely T-shaped; distance between internal pseudostigmatal sockets 97y. (97-130, X = 113); anterior pseudostigmatals sparsely plumose, about 22u, long, arising mesad of pseudostigmatal socket. Venter. All setae elongate, stout, distinctly plumose, 54 to 90y. long; internal ventral I much longer than external; lateral and ventral margins of circumgnatho- somal foramen not greatly thickened, not forming distinct gna- thosomal collar; apodemes II incomplete medially. Hysterosoma. Dorsum. All dorsal setae sparsely plumose, stout, elongate; dorsals I anterior to line drawn between laterals I by about two areolar diameters, 103y- (85-105, X= 99) long, 96y. (96-113, X = 106) apart; laterals I 119, (108-123, X = 118) long, much closer to respective dorsal than dorsals to each other; dorsals II 1244, (124-144, X = 138) apart, 1238p (123-184, X = 128) long; dorsals III extremely elongate, 161ly. (157-189, X = 169) long, only 59u. (59-66, X = 64) apart, laterals III 8ly (81-90, X = 86) long, arising slightly anterior to and well laterad of dorsals; dorsals IV 138y. (138-175, X = 145) long, about 40 to 50y apart; laterals IV 59u. (59-74, X = 63) long, distinctly more slender than dorsals and arising slightly anterior to and well laterad of them; setae of seg- ment V aciculate, sparsely plumose, arising from the ventral margin of the opisthosoma; internal caudal as stout (or stouter) at the base than dorsals III and IV, 64y. (58-70, X = 65) long, external caudals I and II subequal in size, 19 to 28y. long, distinctly less than half as wide at base as internal caudals, external caudals I arising much closer to internal caudals than to external caudals II; posterior genital sclerite indistinct, median genital sclerite single, rounded (to irregular) in shape. Venter. All setae of posterior ventral plate elongate, stout, sparsely plumose, about 55 to 85y. long; line drawn between first axillaries dividing distance between internal and external presternals approximately in half; external presternals laterad of internal presternals, much closer to them than to first axillaries; line drawn between second axillary and internal pre- sternal passing barely anterior to areolus of external presternal, external poststernals arising behind line drawn between _post- eromesal margins of coxal foramina IV by about a third the length of the coxal base; apodemes III indistinct (to weakly distinct as far as internal presternals); apodemes IV weak but distinct, bending posteriorly at the sides to join mesal margins of coxal foramina III behind their centers. THe Famity PYEMOTIDAE OS Legs (all measurements in y.). Length. Leg I, 77 (77-81, X= 78); leg II, 123 (123-147, X = 134); leg III, 141 (141-162, KX = 149); leg IV, 238 (233-289, X = 261). Width. Leg I, 18 (18-20, x 19) lee Wl Os (25-31) X28) leg Mill 2386(23-20, Xi — 24) leg IV, 18 (18-25, X = 28). Segment lengths. Tr I, —?? (17-28, X = 20); tita I, 46 (44-46, X = 45); tr II, — *° (36-45, X = 42); ta II, — *° (48-58, KX = 54); tr III, — °° (42-55, X = 48); ti III, 24 (24- 23 = 26) tails Gil (52-61, Xx — oe ex IV, 59 (59-74 xX = 6o))- tr IV, 34 (34-44, X = 40); fege IV, 14 (X = 14); ta IV, 102 (102- 128, K=115). Setation. Leg I: cx 1T, tr 2T +1 modified; fege AT, tita 11T + 9S; leg II: cx IT, tr 8T, fege 8T, ti 4T + 2S, ta 6T; leo ie cx EE tr 2 fece) 20 tr 4 1S ta Gi leg IV: ex Ie tr 21, fege 11, ti 47 + IS, ta 6D. Tip of seta c of trochanter I not proflexed, seta d similar in shape but longer and more slender; solenidium 3 of tibiotarsus I about 14, thickly rodlike, attenuate, apex sharply acuminate; solenidium 4 nearly as long as 3, similar in shape but thinner, apex less acumi- nate; length of pinnaculum distinctly more than twice its width; tibiotarsus I distinctly wider than femurogenu I, elliptical; tibial solenidia of legs II and III basal, clavate, short; solenidium of tarsus IT about 11, bullet-shaped, apex sharply acuminate; coxa IV rectangulate, ratio of length to width, 1:1.8 to 1:2.1; lateral setae of trochanter IV and tibia IV greatly elongate, the former reaching nearly to, the latter well beyond claw IV; tarsus IV greatly elongate, 102u. (102-128, X = 115) long, evenly attenuate, the pretarsus also elongate and slender, about 22u; claws IV greatly reduced, nearly vestigial; ratio of femurogenu IV to tarsus IV, 1:7.3 to 1:9.2. Hasits. Nothing is known of the habits of this species. DIsTRIBUTION. Known only from the type locality. Type MatertaL. Female holotype and four female paratypes from Volo, Lake County, Illinois, October 14, 1956, B. Greenberg, from sphagnum moss in tamarack bog (berlese). Typr Reposirortes. Holotype and one paratype (no. 1) in the Chicago Natural History Museum, Chicago, Illinois, one paratype (no. 2) in the Snow Entomological Museum at The University of Kansas, one paratype (no. 3) in the U.S. National Museum, Wash- ington, D. C., and one paratype in the Zoologisches Institut, Uni- versitat Erlangen, Erlangen, Germany. 33. This measurement not possible from holotype. 214 THe UNIVERSITY SCIENCE BULLETIN Acinogaster, new genus (Figs. 90-92 ) Type, Acinogaster marianae, new species. This new genus includes six previously undescribed forms as- sociated with doryline ants in both the Old and New Worlds. Its members are very closely related to and considered by this author to form an “exgroup” of the genus Parapygmephorus. They are easily separable from all members of that genus by several interest- ing specializations as noted below. Only the non-gravid females are known. Diacnosis. Females are separable from those of all other genera in the tribe in that at least internal ventrals I and II and axillaries I and II are modified, being enlarged, flattened, and of varying shapes but not setiform. DESCRIPTION OF NON-GRAVID FEMALE. Length, 170 to 300y; width, 90 to 175y; distance between anterior sternocoxal condyles of coxae III, 70 to 120u.; hysterosoma elliptical to ovate; integument of body weakly to moderately sclerotized, lacking engravings; striae absent on most lateral portions of posterior ventral plate; body with or without distinct, coarse punctures. Gnathosoma. As wide as to slightly wider than long; palpal solenidium 2 large, globular or capitate; esophagus straight, thick, sometimes with short thickened area, “bow-tie” structures lacking. Propodosoma. Integument weakly sclerotized, similar to that of hysterosoma or with heavily sclerotized areas which are variable in shape and extent. Dorsum. Inversely T-shaped, usually thickly so; distance between internal pseudostigmatal sockets distinctly greater than length of propodosoma; lateral margins of cireumgnathosomal foramen not greatly thickened to form foraminal collar; posterior margin to posterior third covered by first hysterosomal tergum; pseudostigmata located between posterior half and posterior quar- ter; peritremes circular, large, weakly rimmed or rim lacking, their mesal margins emarginate or not; peritremes widely separated, located in anterior third of dorsum or in propodosomal shoulders; esophageal “finned” structures indistinct. Venter. At least three pairs of setae of antericr ventral plate modified bladelike, paddle- like, or clublike; internal ventrals I anterior to, in line with, or posterior to external ventrals I; internal ventrals II usually in trans- verse line with, rarely well anterior to external ventrals II, often as close to anterior median apodeme as to latter setae; external ventrals THE Famity PYEMOTIDAE 215 I not furcate; coxa [ articulated obliquely to longitudinal axis of body; apodemes II variable but always weak, appearing as integu- mental creases or hidden within strongly sclerotized venter, arcuate or sharply U-shaped, complete or not; secondary suture variously developed, also usually appearing creaselike; anterior median apodeme complete, variable in distinctness; posterior marginal apodemes usually straight, often greatly thickened at lateral angles, thickening rarely extending bandlike mesally across venter; pleated conjunctival band distinct between propodosoma and hysterosoma or not. Hysterosoma, Dorsum. First segment overlapping posterior margin to posterior fourth of propodosoma; fifth segment bearing three pairs of apicoventral setae; vulva apical, distinct, apical por- tion of vaginal wall usually sclerotized but never thickly so; posterior genital sclerite usually elongate and evenly attenuate posteriorly, flaplike but often indistinctly so, or occasionally broadly triangular and flaplike; median genital sclerite present. Venter. Posterior ventral plate with six pairs of setae, its hind margin undulate or broadly emarginate; first axillaries never anterior to anteriormost margin of coxal foramina III by more than twice their areolar di- ameters; posterior ventral plate distinctly produced over anterior portion of coxae III; external presternals always arising well anterior to apodemes IV, the distance between them much greater than that between internal presternals; external poststernals always arising less than half (usually less than a quarter) the basal width of coxa IV behind mesal, posterior sclerotization of fourth coxal foramen; apodemes III indistinct to distinct, if distinct reaching laterally as far as middle of coxal foramina III but not bending posteriorly to join it; apodemes IV distinct or not, if so then bending posteriorly to join coxal foramina III at or behind middle, joining posterior median apodeme on or between lines drawn between anterior third and posterior third of coxal foramina III; posterior median apodeme distinct or not, variable, if distinct, portion behind apodemes IV usually smaller than that anterior to apodemes IV; well defined band of opisthosomal pleats absent behind coxae IV and dorsally between segments III and IV; opisthosoma distinctly longer than distance between anterior margin of coxa III and posterior margin of coxa IV; opisthosomal venter lacking integumental engravings. Legs. Moderately elongate, stout, ratio of length to width of trochanter IV, 1:1.4 to 1:2.2; legs I subequal, to slightly or distinctly shorter than legs II, always much longer than gnathosoma; tibio- 216 THe UNIversIty SCIENCE BULLETIN tarsus I two to two and one-half times as long as wide, as wide as or only slightly wider than femurogenu I; pinnacula reduced to moderately developed, height sometimes one and one-half times apical width; mediolateral sclerotized area of tibiotarsus I usually well developed, becoming enlarged and circular or nearly so dorsally; claw I dilate basally or not, pedicellate, pedicel usually longer than wide, constricted basally, thumb weakly developed; trechanter I with three setae, the apical portion of seta c weakly or strongly proflexed; tarsi II and III each with six tactiles; basal solenidium of tarsus II displaced basad, arising from apical margin of tibia II; basal solenidia of tibiae II and III present; pretarsi II to IV short, never more than one-third the length of the tarsi; claws II and III with large basal dilations, claw IV simple, smaller or rarely absent; pulvilli II and III always distinctly longer than claws, pulvillus TV variable in length; coxa III elongate-subcordate, dis- tance between anterior and posterior sternocoxal articulations, 20 to 30u.; coxa IV rectangulate to short-rectangulate, basal third not or only slightly constricted, basal fifth not or only slightly bulbose, ratio of length to width approximately 1:1.6 to 1:2.0; trochanter IV usually distinctly shorter than, rarely subequal to coxa IV; femuro- genu IV usually wider than long, bearing one seta; tarsus ITV slender and elongate, evenly attenuate, always at least twice as long as tarsus III; ratio of femurogenu IV to tarsus IV, 1:4.5 to 1:6.8. Hasirs. Little is known of members of the group. The non- gravid females are common as phoretics upon many species of army ants, usually between coxae I. According to C. Rettenmeyer (in litt.) the number present per ant is small. To date, six species, all undescribed, are known, these found upon fourteen species of Dorylinae. Table VII summarizes our knowl- edge of the host relationships of the various American species of the genus. Although sample sizes are variable, this table presents some useful information as regards host relationships. Only one of the six species of Acinogaster is restricted to a single species of ant (sp. V, to Neivamyrmex legionis), while each of four species of ants is seen to carry but one mite species as follows: Eciton hamatum with only marianae, E. vagans with only species XX, and Neivamyrmex opacithorax and nigrescens with only species XXX. Species XX is apparently most catholic in its host preference, being known from 11 different species of ants. Members of the genus offer opportunity to study interrelationships between geographical and host variation. THe FamMity PYEMOTIDAE 217 Tas_eE VII.—Distribution of Acinogaster sp. Upon Various American Doryline Ants Percent Percent, Number mite total Mite species Ant species mites/ant | species/ no. all species ant mites/ant species species * A. marianae..| Eciton hamatwm (Fabricus) 53 17 100 E. burchelli (Westwood)... . 255 83 95 100 A.sp IIl....| Heiton dulcius Forel....... 4 4 44 Nomamyrmex hartigi....... 1 1 25 (Westwood) N. esenbecki (F. Smith)..... 2 2 33 Labidus praedator (F. Smith) 84 81 60 L. coecus (Latreille)....... 5 5 71 Newamyrmex pilosus....... 5 5 71 (F. Smith) N. gibbatus Borgmeier...... 1 1 33 N. legionis (F. Smith)...... 1 1 14 100 A.sp. XX....| Heiton vagans (Olivier)..... 3 3 100 LET CLULCTUS | ee epee a ee eae 5 5 56 E. mexicanum Roger....... 11 11 100 BieOURCNelliee mie aan 14 13 5) Nomamyrmex hartigi....... 3 3 75 INES ONG MOCO We a Saeco a ois one 4 4 66 Labidus coecus............ 2 2 ID GUACMOR SF 5's & 6s oc ak o7 54 40 Newamyrmex pilosus....... 2 2 29 INE QUDDOLUSe a eae ee ee 2 2 66 ING WATHOMIS Son k 628 Sisk ee tes D 2 29 100 A.sp. XXX? | Newvamyrmex nigrescens... . 13 46 100 (Cresson) N. opacithorax (Emery)... . 15 54 100 100 ZAMS DSi Vine ean Newamyrmex legionis...... 4 100 57 a. Obtained by totalling number of mites for each ant species, apportioning each mite species therefrom. b. May be only geographic variant of XX. DistRiBuTION, Various species are known from Brazil, British Guiana, Trinidad, Panama, Kansas (U.S. A.), and Angola. SPECIMENS EXAMINED. In addition to the type species, as follows: Species II, Ancoita: Nova Lisboa Prov., 199° ° from Dorylus (Anomma) sp. Species III, PanamMA (C.Z.): Barro Colorado Is., 84 9 9 from Labidus praedator. Same,5 9 ¢@ from L. coecus. 218 THe UNIVERSITY SCIENCE BULLETIN Same, 42 @ from Eciton dulcius. Same, 22° ¢ from Noma- myrmex esenbecki. Same, 1¢@ from Neivamyrmex gibbatus. Same, 5° 9 from N. pilosus, Cerro Galera (K-6 Road), 19 from Nomamyrmex hartigi.. Brazit: Nova Teutonia, 27° 11’S., 52°28’W., 1¢ from Neivamyrmex legionis. Species XX. Panama (C.Z.): Barro Colorado Is., 39 9 from Eciton vagans. Same, 59 9 from E. dulcius. Same,11 9? 9 from E. mexicanum. Same,4¢@ 9 from Nomamyrmex esenbecki. Same, 29 2 from Labidus coecus. Same, 572 @ from L. praedator. Same, 29 ¢ from Neiva- myrmex pilosus. Same, 2? @ from N. gibbatus. Cerro Galera (K-6 Road), 39 @ from Nomamyrmex hartigi. Braz: Nova Teutonia, 29 9° from Neivamyrmex legionis. British GUIANA: Oronoque River at New River, 109° 9 from Eciton burchelli cupiens. Species XXX. Kansas: Lawrence, 15? ? from Neiwva- myrmex opacithorax. Univ. Kansas Nat. Hist. Res., 122 @ from N. nigrescens. Topeka, 19 from N. nigrescens. Species V. Brazit: Nova Teutonia, 27°11’ S., 52°23’ W., 49 9 from Neiva- myrmex legionis. Acinogaster marianae,** new species DEscRIPTION OF NON-GRAVID FEMALE. Length, 333, (267-346, X = 31140 =S_E. 8.14); width, 183y (157-210, X = 181.70 =s2u; 5.91); width between anterior sternocoxal condyles III, 126y (101-126, X = 115.44); body moderately sclerotized. Gnathosoma. Length, 26y. (24-30, X = 26.50); width, 25y, (25-31, X = 28.50 + S.E. 0.65); dorsals stout, internals slightly longer than externals, arising well mesad and slightly in front of them; distance between externals greater than that between external and respective lateral; palpal tooth large, distinct; solenidium II large, globular or capitate, solenidium I strobiloid, arising nearly laterad Ota. Propodosoma. Dorsum. Broadly inversely T-shaped (shape vary- ing somewhat according to mount), distance between internal pseudostigmatal sockets, 91y. (72-91, X = 84.33); sclerotization dis- tinctive, uneven, heaviest in circular area immediately mesad of each pseudostigma, this area forming a broad tubercle; posterior margin of dorsum excavated, thickly sclerotized, sclerotized areas becoming broader at the sides; anterior pseudostigmatal setae short, nude, bristlelike, posterior pseudostigmatals stout, sparsely plu- mose, arising well laterad of and only slightly posterior to internal 34. This species is named in honor of Mrs. Marian E. Rettenmeyer who, with her husband Carl, collected the type series. THe FamMiILty PYEMOTIDAE 219 pseudostigmatal socket. Venter. External and internal ventrals I and internal ventrals IJ modified, greatly thickened and flattened; internal sternals I and II clavate, similar in shape, but I distinctly larger than II; external ventral I broader than internal but elongate- cylindrical, its apex acuminate; external ventral II stoutly setiform, plumose; internal ventral I slightly behind line drawn between external ventrals I (but relationships depend greatly upon mount); internal ventrals II approximately in transverse line with external ventrals II; posterior marginal apodemes well developed, becoming greatly enlarged at the sides; anterior median apodeme prolonged taillike behind union with posterior marginal apodemes; band of pleated conjunctiva distinct (or indistinct) dorsally between pro- podosoma and hysterosoma. Hysterosoma, Dorsum. All dorsal setae stout, setiform, elongate, moderately plumose; dorsals I 56u. (51-64, X = 60.89) long, 63u. (53-69, X = 61.90 + S. E. 1.84) apart, distinctly closer to respective laterals than to each other; laterals I arising only slightly behind line drawn between dorsals, 99y. (78-108, X = 92.78) long; dorsals ou. (46-5605 X— 52-70 (5B 125) longs Sin (59-3855 x = 76.50 = S. E. 2.71) apart; dorsals IIT 68». (59-76, X = 67.22) long, 74. (59-84, X = 72.80 + S. E. 2.44) apart; laterals III well laterad of and slightly behind dorsals III, 44y. (31-50, X = 43.44) long; dorsals IV 88y (70-92, X = 83.20 = S. E. 2.35) long, 62y, (45-76, X = 61.30 + S. E. 3.30) apart; laterals IV posterolateral to dorsals IV, 50. (40-51, X = 46.30 + S. E. 1.34) long; internal caudals much longer and stouter than other setae of segment V, 44y. (36-45, X = 40.90 = S.E. 0.96) long; areoli of external caudals I nearly con- tiguous with those of internal caudals, 17y. (15-18, X = 16.57) long; external caudals II slightly longer than external caudals I and of the same conformation, 20y. (17-23, X = 19.00 + S. E. 0.64), about three to four times as far from external caudals I as the latter are from internal caudals; median genital sclerite distinct, globular (to irregular) in shape. Venter. First and second axillary setae modi- fied, remaining setae of posterior ventral plate setiform, moderately and subequally sized, nude or nearly so, flagellate; first axillary bladelike, straight, with distinct prickle, second axillary similar but moderately proflexed near tip; external presternals well in front of apodemes IV, only slightly behind first axillaries, in lateral one quarter of posterior ventral plate; internal poststernals slightly laterad of internal presternals, approximately on transverse line drawn across posterior termination of posterior median apodeme; 22.0, THe UNIVERSITY SCIENCE BULLETIN poststernals not reaching halfway to posteriormost margin of opis- thosoma; mesal portion of hind margin of posterior ventral plate broadly and shallowly lobate, lateral margins indistinct. Legs (all measurements in y.). Length. Leg I, 97 (84-100, K = 94:80 22S. E: 1:67); lee Il, 121 (100-126, Xk = 117.110 22859 Beane leg ITI, 128 (109-140, X —126.67); leg IV, 208 (173-214, X = 200.70 +S.E. 4.38). Width. Leg I, 14 (14-20, X = 17.60 + 0.65); leg IL, 25 (23-80, X = 26.90 = S.E.. 0.82): leg IM, 22; (20225) x Daal + §. E. 0.53); leg IV, 18 (15-19, X = 17.30 +S. E. 0.42). Segment lengths, Tr 1, 26 (25-30, X = 26.33). tita 1) 54 (46547 x — 550 +S. E. 0.85); tr Il, 41 (80-41, X = 37.00 =S.E. 1.21). ta I, 44 (39-48; X == 4490 = SE. 1.038) tr I, 47 (37-5 xX 488) III, 23 (20-23, X = 21.80 = 0.33); ta III, 47 (40-52; K = 45.40 1.16): cx 1V,56 (44-57; X = 53.70 = SB: 146). ING 84 le are X = 34.50 + S_E. 0.69); fege IV, 13 (11-18, X = 1230+ S. EF. 0.21). ta IV, 81 (66-85, X —78.10+S.E. 1.98); width of coxa IV, 26 (23-30, X = 27.00). _Setation. leg 1: ex IT, tr 39 fege 41, tita 11T + 8S; leg II: cx IT, tr 8T, fege 3T, ti 4T + 2S, ta 6T; leg Ill; ex 1T, tr 2T, fege 27, ti 47 + 1S, ta 6T: lee 1V.-cx ie ty 2 fege 1T, ti 4T + 1S, ta 6T. Apex of seta c of trochanter [| distinctly proflexed, seta d sinuate, resembling sense seta; mediolateral sclerotized plate of tita I prominent between solenidia 2 and 3, marginal, rounded; solen- idium 3 of tibiotarsus I short, strobiloid, distad of above spot, solenidia ] and 2 arising together beneath it, one clavate-pedicellate, the other bullet-shaped; claw I moderately sized, arising from funnel-shaped pedicel; solenidium of tibia II basal, strobiloid, solenidium of tarsus IT displaced basad, arising from apical margin of tibia II, bullet-shaped, with apical prickle; solenidium of tibia III distinctly smaller than that of tibia II, clavate; coxa IV rectangu- late, ratio of length to width, 1:1.5 (1:1.4 to 1:1.6); external apical seta of tibia IV longest seta of leg; tarsus IV very elongate, slender, tapering basally; ratio of femurogenu IV to tarsus IV, 1:3.4 (1:3.4 to 1:4.7). Hasirs. See description of genus. DisTRIBUTION. Known only from Panama and from Trinidad, British West Indies. Type MatertaL. Female holotype from Barro Colorado Island, Panama (Canal Zone), June 16, 1956, C. W. and M. E. Retten- meyer, from Eciton hamatum, colony E-189 (Slide No. 1804). Two THe FAMILY PYEMOTIDAE 221 paratypes (nos. 3 and 4) same data except taken August 17, 1956, from E. burchelli, colony No. E-253. Two paratypes (nos. 6 and 8) same data as preceding except taken on May 12, 1956, from colony No. E-148. Paratype No. 8 same data except taken April 20, 1956, from colony No. E-149. Two paratypes (nos. 1 and 5) same data except taken March 1, 1955, C. W. Rettenmeyer from colony 55-B-II. Two paratypes (nos. 2 and 7) from 7 mi. N. W. of Balboa, Cocolli, Panama (C.Z.), January 1, 1957, W. Lundy, from E. burchelli. Tyre Repositories. Holotype and four paratypes (nos. 3, 4, 8, 9) in the Snow Entomological Museum at The University of Kansas. One paratype (no. 2) in the Zoologische Institut, Universitat Erlangen, Erlangen, Germany. One paratype (no. 1) in the U.S. National Museum, Washington, D.C. One paratype in the British Museum (Natural History), London. One paratype (no. 6) in the private collection of C. W. Rettenmeyer, Kansas State University. One paratype (no. 5) in the private collection of the author. SPECIMENS EXAMINED. Other than the type series, as follows: PanaMA (C.Z.): Barro Colorado Island, various dates between Feb. 9, 1955, and Aug. 17, 1956, C. W. and M. E. Rettenmeyer or C. W. Rettenmeyer, 247° 9? on Eciton burchelli. Same locality and collectors, various dates between March 2] and July 9, 1956, 539 ¢ from E. hamatum. Seven mi. N. W. of Balboa, Cocoli, Jan. 1, 1957, W. Lundy, 69 ¢ from E. burchelli. Trremap: Arima Valley, Spring Hill Estate, April 6, 1935, N. A. Weber, 1 9 from E. burchelli. South margin, Nariva Swamp, March 10, 1935, N. A. Weber, 1 ¢ from E. b. wrichi Forel. Northern Range, just north of St. Augustine, June 2, 1935, N. Weber, 19 from E. b. urichi. BritisH GuraANA: Oronoque River at New River, July 21, 1936, N. A. Weber, 39 9 from E. b. cupiens Santschi. Pseudopygmephorus, new genus (Figs. 93-94 ) Type, Pygmephorus tarsalis Hirst, 1921. This genus is erected to contain a number of forms closely related to and perhaps directly ancestral to the genera Microdispus and Neopygmephorus. Primitive forms of the genus, e. g. P. sellnicki (Krezal, 1958) superficially resemble members of Siteroptes (Sit- eroptoides ), but closer examination indicates that they are probably not direct descendents of the latter, but are more closely related to Pygmephorus and certain Parapygmephorus, and the development 9—1367 229, THe UNIVERSITY SCIENCE BULLETIN of such characteristics as elongate legs I and sicklelike claws on legs I to III may be regarded as secondary. Members of the genus show increasing specialization, forms considered by the author to be the most advanced closely resembling less specialized forms of the genus Neopygmephorus. Because the members of the two groups may be satisfactorily separated, and because of the size of the genus Neopygmephorus, the author prefers to consider Pseudo- pygmephorus a distinct genus containing at least fifteen described forms. Dracnosis. In addition to the key characters given above, females may be distinguished from those of other neopygmephorines by the following combination of characters: external presternals arising from or barely in front of apodemes IV or, if not, then apodemes III distinct laterally and bending posteriorly to meet anterior sterno- coxal condyles III; claw I rarely lacking, otherwise moderate to small in size, without thumb, always arising from a distinct, stalk- like pedicel. DESCRIPTION OF NON-GRAVID FEMALE. Length, 160 to 300y; width 85 to 160u; width between anterior sternocoxal condyles of coxae III, 50 to 100u; hysterosoma rather narrowly to broadly elliptical in dorsoventral aspect, posterior margin evenly rounded to truncate; body weakly to moderately sclerotized, without scalelike or “peb- bled” integumental markings, not striate on most lateral portions of posterior ventral plate but usually distinctly punctate. Gnathosoma. Always distinctly longer than wide; palps semipal- mate, possessing one or more apical teeth, comprising about half length of gnathosoma; chelicerae small, usually indistinct; small circumesophageal pump sometimes present, a second, “bow-tie- shaped” circumesophageal structure often present behind pump. Propodosoma. Dorsum. Elongate, inversely T-shaped or trap- ezoidal, length usually much greater than, rarely subequal to dis- tance between internal pseudostigmatal sockets; anterior portion prolonged hoodlike anteriorly and ventrally over gnathosoma, cir- cumgnathosomal foramen ventral; posterior margin narrowly cov- ered by first hysterosomal segment; pseudostigmata located in posterior one third; two distinct pairs of propodosomal setae; peritremes dorsal or dorsolateral, elongate and throughlike, long- guttate, or guttate, if as first two then diverging anteriorly, if as the latter then stigmatal openings in median or posteromedian excavation; esophageal “finned” structure distinct or not, if distinct, then posterior globose structure always more prominent than THe FAMILY PYEMOTIDAE 293 elongate anterior one, fins indistinct in both. Venter. Internal ventrals I distinctly anterior to externals, internal ventrals II slightly to well anterior to external ventrals II, usually much closer to them than to anterior median apodeme, both pairs rarely approximate to coxal foramina II; external ventrals I commonly bifurcate; coxa I articulated at nearly a right angle to longitudinal axis of body or distinctly obliquely; apodemes II arcuate, together sometimes lyrate, or weakly W-shaped, always complete; secondary transverse apodeme usually absent, rarely present laterally; anterior median apodeme strong and complete; posterior marginal apodemes strong, straight to gently arcuate; distinctly pleated conjunctival band absent between propodosoma and hysterosoma. Hysterosoma, Dorsum. First segment covering only posterior quarter to posterior third of propodosoma; fifth segment bearing two (rarely) or three pairs of ventral to apicoventral setae; vulva apical and distinct, often located within genital sulcus; vaginal walls usually distinct apically, globose or elliptical pumplike en- largements usually lacking (present in sellnicki group); posterior genital sclerite irregularly rounded, triangular, or rarely, tri- angular and flaplike. Venter. Posterior ventral plate with five or six pairs of setae, if five then second axillaries missing; hind margin of posterior ventral plate variable, entire and linear or lobate or (rarely, sellnicki group) tripartite, incisions reaching nearly to posterior foraminal margin of coxa IV; external presternals always arising from or barely before apodemes IV, or if well anterior to apodemes IV, then directly behind internal presternals; transverse distance between external presternal and internal presternal as great as or greater than longitudinal distance between external presternal and first axillary; external poststernal behind posterior foraminal thickening by less, usually much less, than basal width of coxa IV; apodemes III variable in definition and extent; apodemes IV directed toward anterior third or quarter of, reaching coxal foramina III or not, if so, then bending sharply posteriorly to join foramina behind their centers (sellnicki group); if apodemes IV reach foraminal margin, then apodemes III are well defined and bend sharply posteriorly to meet anterior sternocoxal condyles; posterior median apodeme well defined, portion behind apodemes IV not more than one and one-half times that anterior to apodemes IV; bands of opisthosomal conjunctival pleats absent; opisthosoma distinctly longer than distance between anterior margin of coxal foramen III and posterior margin of coxal foramen IV; opisthosomal venter lacking integumental engravings. 224 THE UNIVERSITY SCIENCE BULLETIN Legs. Leg I usually longer than or subequal to, always wider than leg II, tibiotarsus I elongate, at least twice as long as wide, only rarely distinctly wider than femurogenu I; pinnacula lacking; mediolateral sensory area present or not, if present marginal and irregular in shape; claw I simple or basally dilated, small to moderate in size, distinctly pedicellate, pedicel always much longer than wide, stalklike, thumb lacking or greatly reduced; trochanter I with three setae, seta c with apical portion proflexed; tarsi II and III each with six tactiles; solenidia of tibiae II and III present; pretarsi II to IV short, one-quarter the length of the tarsi or less; claws II to III simple or dilated basally, claws IV simple, subequal to or smaller than those of II and III; coxa III subcordate or tri- angular as viewed in ventral aspect; coxa IV short-rectangulate to rectangulate, basal third with or without distinct constriction, basal fifth distinctly bulbose or not, ratio of length to width, 1:1.3 to 1:2.2; trochanter IV at least slightly shorter than coxa IV; tarsus IV short to moderately long, usually abruptly narrowed, always dis- tinctly longer than tarsus III; ratio of femurogenu IV to tarsus IV, 12:8 to l:357, Species thought to belong to Psewdopygmephorus: abdominalis ( Berlese, 1904) (Pygmephorus ) abdominalis polonicus (Willmann, 1949) (Pygmephorus) abdominalis var. sudetica (Willmann, 1956) (Pygmephorus) arcticus (Sig Thor, 1934) (Pediculoides ) arvorum (Krezal, 1959) (Pygmephorus ) decumanus (Krezal, 1959) (Pygmephorus ) disparisetae (Jacot, 1936) (Pygmephorus ) haarloevi (Krezal, 1958) (Pygmephorus ) inconspicuus ( Berlese, 1904) (Pygmephorus ) kochi (Krezal, 1959) (Pygmephorus) ?quadratus (Ewing, 1917) (Pygmephorus ) = (sellnicki Krczal? ) sellnicki (Krezal, 1958) (Pygmephorus ) suecicus (Krezal, 1959) (Pygmephorus ) tarsalis (Hirst, 1921) (Pygmephorus ) togatus (Willmann, 1942) (Pygmephorus ) willmanni (Krezal, 1959) (Pygmephorus ) Hasits. Unknown. None have been found to be phoretic upon insects, although two species, haarloevi and willmanni are known only from the nests of the ant, Lasius flavus Mayr. Species known only from decaying plant materials are kochi, decumanus, and ab- dominalis var. sudetica. Other forms have been taken from soil and from small mammals. THe FamMiIty PYEMOTIDAE 295 While some forms such as haarloevi and willmanni may be re- stricted in their host associations, the ecological distribution of some species is amazingly broad. P. sellnicki, for example, is recorded by Krezal, (1959b), from various small rodents and insectivores, from soil, and from lily bulbs. The present author finds a form considered to be the same species in various other habitats (see “Specimens examined’ ). Mites associated with rodents are often found in the nests or in food caches, and, as Vitzthum (1943), has stated, probably stand only in a phoretic relationship to them. SPECIMENS EXAMINED. Five species, including sellnicki and probably tarsalis, from most of the United States and from England and Japan, from the following situations or hosts: commercial mush- room houses, peach orchard soil, grass, grain, leaf mold, moss, decayed tomatoes, pea ensilage, frostscar, oat straw, lily bulbs, African marigolds, manure, decaying log, treehole of Fagus grandi- folia, unoccupied nest of Microtus sp., from Peromyscus maniculatus gracilis (LeConte), burrow of Cynomys sp. and debris pile of Neo- toma sp. (all Cricetidae), from Pipistrellus subflavus (F. Cuvier) (Chiroptera ). Neopygmephorus, new genus Type, Pygmephorus blumentritti Krezal, 1959. The author erects this genus to include a large number of forms placed by previous authors in the genus Pygmephorus. As presently conceived, the genus remains one of the largest in the family, con- taining at least thirty species, these divided into the two subgenera Allopygmephorus and Neopygmephorus sensu stricto. The former contains but two forms which differ greatly from Neopygmephorus s. str., and secondarily resemble some Microdispini in having lost one pair of propodosomal and one pair of caudal setae and because the most lateral margins of the posterior ventral plate possess distinct pleats. Because intergradation between these more spe- cialized forms and typical species of Neopygmephorus seems to oc- cur, they are considered as subgenera herein. Species of Neopygmephorus are most closely related to those of Parapygmephorus and Pseudopygmephorus. Except for the male of one species, to be described elsewhere, only the non-gravid females are known. Species are known from Europe, Asia, North and South America, Australia, and Antarctica. 296 THe UNIVERSITY SCIENCE BULLETIN Diacnosis. Females of Neopygmephorus are separable from those of all other genera in the tribe by the following combination of characters: leg I usually more slender than, rarely as wide as leg II; coxa IV usually broadly constricted near basal third, coxal base appearing bulbose or bilobate, or, if not, then legs long and slender, ratio of length to width of trochanter IIT 1:2.3 to 1:4.2. DEscrIPTION OF NON-GRAVID FEMALE. Length, 145 to 390y; width, 70 to 230u; distance between anterior sternocoxal condyles III, 70 to 155u.; hysterosoma rather narrowly to broadly ellipitical or oval, body moderately to heavily sclerotized, only rarely with integu- mental engravings, these restricted to the lateral or posterolateral portions of the posterior ventral plate if present; striae rarely present on most lateral portions of above plate; body usually distinctly punctate. Gnathosoma. Well developed, usually only slightly longer than wide, never wider than long; chelicerae tiny, if distinct, needlelike. Propodosoma. Dorsum. Inversely T-shaped to broadly trape- zoidal, shape greatly dependent upon mount; lateral margins of circumgnathosomal foramen not thickened or enlarged to form collar; dorsum narrowly to nearly completely covered by first hysterosomal segment; pseudostigmata usually located in posterior third; one (rarely) or two pairs of propodosomal setae; peritremes usually rimmed, circular or transversely ovoid, their mesal margins nearly always broken or emarginate; peritremes usually smaller than, rarely subequal to diameter of external pseudostigmatal socket; peritremes located in mesal or posterior (rarely) third of propodosoma; esophageal “finned” structures distinct or not, if distinct then usually three in number. Venter. Internal ventrals I usually posterior to or in a transverse line with external ventrals I, rarely anterior to them; internal ventrals II distinctly anterior to externals I] and much closer to them than to anterior median apodeme; external ventrals I forked or not; coxa I always articu- lated distinctly obliquely to longitudinal axis of body; apodemes II usually appearing arcuate and complete (see morphology p. 46), terminating anteriorly at external sternal [ or incomplete laterally; secondary transverse apodeme usually present and distinct as far as apodemes II (see morphology, p. 46), or more rarely distinct only laterally or absent entirely; anterior median apodeme com- plete or not; posterior marginal apodemes moderately developed medially, straight or gently arcuate; distance between apodemes I THe Famity PYEMOTIDAE aN and II distinctly less than that between apodemes IT and posterior marginal apodemes; pleated conjunctival band not visible between propodosoma and hysterosoma. Hysterosoma. Dorsum. First tergum usually covering only pos- terior quarter, rarely posterior three-quarters of propodosomal dorsum; fifth segment bearing two or (usually) three pairs of setae, these apical or apicoventral; vulva apical, distinct, vagina some- times sclerotized apically but lacking globose or elliptical pump- like organ; posterior genital sclerite usually distinct, narrowly to broadly triangular and flaplike or (rarely) formed otherwise or absent; median genital sclerite lacking. Venter. Posterior ventral plate with six pairs of setae, its hind margin entire, straight to broadly rounded; external presternals usually arising at least one areolar width anterior to apodemes IV, rarely less; external post- sternals usually arising at least half the basal width of coxa IV behind posteromesal sclerotization of fourth coxal foramen; apo- demes III distinct or not, never bending at sides to fuse with anterior sternocoxal condyles III; apodemes IV (except in Allopygmephorus ) extending at least three-quarters of distance to coxal foramina III; apodemes IV usually directed toward or joining coxal foramina at or in front of line drawn between their centers; apodemes IV joining posterior median apodeme on or above line drawn between points demarking anterior third of foraminal margin, posterior median apodeme well defined, portion behind apodemes IV usually at least one and two-thirds times as long as that anterior to apodemes IV; thick bands of opisthosomal conjunctival pleats absent, thin bands occasionally present, especially dorsally between segments three and four; opisthosoma distinctly longer than distance between ante- rior margin of coxal foramen III and posterior margin of coxal foramen IV. Legs. Legs usually elongate and thin, ratio of length to width of trochanter IV, 1:2.4 to 1:4.5, only rarely under 1:2.7; leg I usually shorter and thinner than, rarely subequal to leg II; tibiotarsus I usually elongate, usually at least twice as long as wide, distinctly wider than femurogenu I or not; pinnacula lacking or if present reduced, free portion not much higher than width at apex; medio- lateral sclerotized area present or not, if present varying in shape and location; claw I usually present, dilate basally, small to mod- erate in size, distinctly pedicellate, pedicel usually much longer than wide, occasionally constricted basally and as wide or wider 228 THE UNIVERSITY SCIENCE BULLETIN than long; thumb lacking or greatly reduced; apical portion of seta c of trochanter I strongly or weakly proflexed and attenuate; tarsi II and III each with six tactiles; solenidia of tibiae II and III present, solenidium of tarsus II arising from the latter’s base; claws II and III basally dilate, claws IV subequal to them in size or smaller, simple; coxa III short, subcordate in inner ventral aspect, distance between anterior and posterior sternocoxal articulations, 16 to 33y. (usually 16-25); coxa IV rectangulate, basal third always slightly, usually greatly constricted, basal fifth distinctly to greatly bulbose, ratio of length to width, 1:2.0 to 1:3.1; trochanter IV vari- able in length, distinctly shorter to much longer than coxa IV; tarsus IV short, variously shaped, usually much longer than, rarely sub- equal to tarsus III; ratio of femurogenu IV to tarsus IV, 1:2.0 to G22: Key to the Subgenera of Neopygmephorus (Females) 1(a). With one pair of dorsal propodosomals; two pairs of setae on posterior segment; first hysterosomal tergum covering at least posterior three- fifths of propodosoma (depending upon mount); tarsus IV short, ratio of femurogenu IV to tarsus IV about 1:2.0 Allepygmephorus, new subgenus, p. 228 1(b). With two pairs of dorsal propodosomals; two (rarely) or three pairs of setae on posterior segment; first hysterosomal tergum covering at most posterior half of propodosoma; tarsus IV longer, ratio of femurogenu IV to tarsus IV rarely less than 1:2.5 Neopygmephorus, sensu stricto, p. 230 Allopygmephorus, new subgenus (Figs. 95-96 ) Type, Pygmephorus matthesi Krezal, 1959. This subgenus includes N. (A.) matthesi and at least one closely related, undescribed form** in the author’s collection. All are phoretic upon various Hydrophilidae (Coleoptera). These mites resemble some Microdispini in lacking one pair of propodosomal dorsals and in the loss of a pair of setae from segment V, in often possessing enlarged poststernals, and because the most lateral por- tions of the posterior ventral plate are striate. They differ from the Microdispini and resemble Neopygmephorus in the conforma- tion of the legs, the presence of a modified seta on trochanter I, the arrangement of the setae of the posterior ventral plate, the shape of the propodosomal dorsum, and in the position and shape of the 35. The author has two forms which he considers to be separate species. Both follow Krezal’s description and illustrations of matthesi closely yet exhibit strong setational and minor structural differences. THE Famity PYEMOTIDAE 229 peritremes. Two species placed herein in the subgenus Neopygme- phorus likewise lack a pair of setae on segment V and possess short tarsi IV. They are possibly intermediates between the two sub- genera, Diacnosis. In above key. DerEscrIPTION OF NOoN-GRAvVID FEMALE. Length, 140 to 220u; width not ascertainable; distance between anterior sternocoxal con- dyles of coxae III 70 to 90y; posterior ventral plate without integu- mental engravings but striate on most lateral portions. Propodosoma. Dorsum. Only anterior pseudostigmatal setae present, these arising between peritremes and _pseudostigmata; dorsum nearly to completely covered by first hysterosomal tergum. Venter. Internal ventrals I anterior to or in transverse line with external ventrals I °°; external ventrals I not furcate; secondary transverse suture absent. Hysterosoma. Dorsum. Fifth segment with only two pairs of apicoventral setae. Venter. External presternals not more than one areolar diameter behind line drawn between internal presternal and corresponding first axillary; external poststernal arising less than half the basal width of coxa IV behind posteromedial sclero- tization of fourth coxal foramen; apodemes IV indistinct or extend- ing only about halfway to mesal margins of coxal foramina III; distance along posterior median apodeme between junctions of apodemes III and IV short, less than 13u. Legs. Short and stout; tibiotarsus I less than twice as long as wide; mediolateral, dorsal sclerotized area circular, distinct; claws IV well developed, simple and more slender but nearly as long as claws III; ratio of length to width of trochanter III about 1:1.8 to 1:2.0; ratio of length to width of coxa IV about 1:2.0 to 1:2.2; ratio of length to width of trochanter IV about 1:2.4 to 1:3.0; tarsus IV short and stout, ratio, length femurogenu IV to tarsus IV, 1:2.0 to 1:2.4; trochanter IV distinctly shorter than coxa IV. Hasirs. Nothing is known of the habits of these mites. They are capable, presumably, of spending considerable periods of time under water. The specimens of Paracymus (below) were captured while submerged. 86. Krezal’s drawing of matthesi shows internal sternals I to be posterior to the externals. The propodosoma of this specimen is tipped, however, and it is probable that these setae are in a transverse line. 230. THe UNIvERSITY SCIENCE BULLETIN SPECIMENS EXAMINED. Species No. I. Mussourt: Weldon Springs, 1 9 from Tropisternus blatchleyi D’Orchymont. Species No. 2. Kansas: Muscotah, 29 9 from beneath pronotum of Paracymus sp.** Subgenus Neopygmephorus sensu stricto (Figs. 97-100) Type, Pygmephorus blumentritti Krezal, 1959. The present author places in this subgenus the forms considered by most American workers to be “typical” members of the genus Pygmephorus. The group is large. Krezal (1959b), includes about fifteen species considered herein to belong to Neopygmephorus s. str. Approximately 12 additional species, all undescribed, are in the possession of the present author. Dracnosis. See key above. DEscripTION OF NOoN-GRAvID FEMALE. Length, 212 to 385y; width, 115 to 220u; distance between anterior sternocoxal condyles III, 70 to 150y. Propodosoma. Dorsum. Posterior margin to posterior half cov- ered by first hysterosomal tergum; two pars of propodosomal setae; peritremes located in mesal third of propodosoma. Hysterosoma. Venter. Apodemes IV joining coxal foramina III at anterior third or quarter, if not, then joining posterior median apodeme on or above line drawn between anteriormost margins of coxal foramina III; external poststernals usually behind postero- mesal sclerotization of coxa IV by at least half the basal width of the coxa; distance along posterior median apodeme between junctions of apodemes IIT and TV usually longer than 18u. Legs. Tibiotarsus I elongate, always at least twice as long as wide; ratio, length to width of trochanter III, 1:2.4 to 1:4.1; ratio, length to width of coxa IV, 1:2.0 to 1:3.1; ratio, length to width of trochanter IV, 1:2.5 to 1:4.5; tarsus IV moderately elongate to elongate, 25 to 100y. long, ratio of femurogenu IV to tarsus IV 213: tOn eGo) Species thought to belong to Neopygmephorus s. str. amplus (Krezal, 1959) (Pygmephorus ) arboris (Krezal, 1959) (Pygmephorus ) bawaricus (Krezal, 1959) (Pygmephorus ) blumentritti (Krezal, 1959) (Pygmephorus ) comatus (Krezal, 1959) (Pygmephorus ) 37. Det. F. N. Young, Indiana University. THe Famity PYEMOTIDAE ETE cultratus (Berlese, 1904) (Pygmephorus ) cultratus var. minor (Berlese, 1904) (Pygmephorus ) delanyi (Evans, 1950) (Pygmephorus ) ellipticus (Krezal, 1959) (Pygmephorus ) formosus (Jacot, 1936) (Pygmephorus ) gracilis (Krezal, 1958) (Pygmephorus ) lithobii (Krezal, 1958) (Pygmephorus ) matthesi (Krezal, 1959) (Pygmephorus ) montanus ( Willmann, 1956) (Pygmephorus ) pannonicus (Willmann, 1951) (Pygmephorus ) pratensis (Krezal, 1959) (Pygmephorus ) scutacaroides ( Vitzthum, 1925) ( Microdispodides ) tumidosetosus (Willmann, 1951) (Pygmephorus ) Hasits. Nothing is known of the habits of the species in this group. Their macrohabitats are similar to those of Pseudopygme- phorus. Two species, lithobii and amplus, are known respectively from Lithobius (Chilopoda) and from the nest of a “wild bee.” Other species are found in soil or litter in assocation with decaying plant material or in the nests of small rodents, insectivores, and birds. SPECIMENS EXAMINED. Fourteen species, including blumentritti, from various parts of the United States including Alaska and from West Australia, from the following situations or hosts: humus in mixed white pine and hemlock forest, litter from fallen Tilia americana L.., leaf litter in alder swamp, melly leaf litter and moss (Australia), abandoned nest of field mouse, occupied nest of Peromyscus sp. (Cricetidae), from “mouse nest,” from nest of Microtus sp. (Cricetidae), from nest of Pitymys pinetorum scalop- soides (Aububon and Bachman) (Cricetidae), debris from nest of Neotoma sp. (Cricetidae ), debris in nest of Colaptes cafer (Gmelin ) (Piciformes), from hooded warbler (Parulidae), from Microtus ochrogaster (Wagner) (Cricetidae), from tundra, from soil, moss, “Brachystachium sp., under bark of apple, punky stump, from tideline detritus. 232, Tue UNIVERSITY SCIENCE BULLETIN LITERATURE CITED ALFARO, A. 1946. El acaro Pediculopsis graminum Reuter y el hongo Nigrospora oryzae (Berk. & Br.), en asociacién parasitaria sobre trigos ara- gonenses. Boletin de Patologia Vegetal y Entomologia Agricola de Madrid, 14:321-334. AMERLING, C. i 1861. Die Milbenkrankheit unserer Getreidearten. Lotos, Zeitschrift fiir Naturwissenschaften, 11:24-27. 1862. Ueber die Naturdkonomie der von ihm beobachtungen Milben, inbesondere der Trombidien. Sitzungsberichte Boehmische Gesell- schaft der Wissenschaften (= Ceska Spole¢nost Nauk, Prague), 1862, pp. 54-56. ATHIAS-HENRIOT, C. 1961. Pygmephorus dominguezi, nouveau Pyémotide algérien (Acari- formes, Tarsonemini). Acarologia, 3(4):571-574. Baker, E. W., and G. W. WHARTON. 1952. An introduction to acarology. xii-+ 465 pp. The MacMillan Co., New York. Banks, N. 1905a. A treatise on the Acarina, or mites. Proc. U.S.N.M., 28:1-114. 1905b. Descriptions of some new mites. Proc. Ent. Soc. Washington, 7:133-142. 1915. The Acarina or mites. U.S. D.A. Reports, 108:1-153. Beer, R. E. 1954. A revision of the Tarsonemidae of the western hemisphere (Order Acarina). Univ. Kansas Sci. Bull., 36: 1091-1387. Beer, R. E. and E. A. Cross 1960. A new genus and two new species of scutacarid mites (Acarina) phoretic upon bees. Jour. Kansas Ent. Soc., 33:49-57. BERLESE, A. 1885. Acarorum systematis specimen. Bolletino della Societa Entomo- logica Italiana, 17:125-135. 1886. La Sottofamiglia dei Tarsonemidi. Ibid., 18-334-354. 1895. Acari, Myriapoda, Scorpiones, huscusque in Italia reperta. Fasc. 75, vols. 1-8, Padova. 1897-99. Gli Acari agrarii. Rivista Patologia Vegetale, 6:1-65; 7:312-344. 1904a. Acari nuovi. Manipulus III. Redia, 2: 10-32. 1904b. Diagnosi di alcune nuove specie di Acari Italiani, mirmecofilie liberi. Zoologischer Anzeiger, 27:12-28. 1911. Alcuni acari entomofili nuovi. Redia, 7:183-186. 1925. Gli Insetti. Vol. 2, Vita e Costumi. ix + 992 pp., Societa Edi- trice Libraria, Milano. BicELow, R. S. 1958. Classification and phylogeny. Systematic Zool., 7(2):49-59. THE FamMiIty PYEMOTIDAE 233 BLANC, G. 1912-13. Revue generale de la famille des Tarsonémides. Bulletin de la Société d'Etudes de Vulgarisation de la Zoologie Agricole, Bor- deaux, 11(6):153-163 (Pediculoides), also 12(1):14-18 (Pedicu- lopsis ), and 12(2):41-50, (Tarsonemidae ). Brapsury, O. C. 1926. Some phases of the embryology of the carnation mite. Jour. Elisha Mitchell Scientific Soc., 42:94-98. Brucker, E. A. 1899. Sur Pediculoides ventricosus Newport. Compte Rendus Hebdo- maires des Seances et Memoires de la Société de Biologie (12)9: 953-955. 1901. Monographie de Pediculoides ventricosus Newport et theorie des pieces buccales des acariens. Bulletin Scientifique de la France et de la Belgique, (6)35(4):365-451. CANESTRINI, G. 1888. Prospetto dell’Acarofauna Italiana. Vol. 3, famiglie Tarsonemini e Tiroglifini e appendice agli Analgesini, pp. 309-425. 1891. Abbozzo del Sistema Acarologico. Atti del Reale Instituto Veneto di Scienze, Lettere, ed Arti, (7)2(1):699-725. CANEsTRINI, G. and A. BERLESE 1884. Sopra alcune nuove specie di Acari Italiana. Atti della Societa Veneto-Trentino di Scienze Naturali, 9:175-182. CANESTRINI, G. and F. FANZAGO 1876. Nuovi Acari Italiani. Ibid., 5:99-111 and 130-142. CANESTRINI, R. 1880. Contribuzione allo studio degli Acari Parassiti degli Insetti. Ibid. vol. 7:154-178. Cooper, K. W. 1937. Reproductive behaviour and haploid parthenogenesis in the grass mite Pediculopsis graminum (Reuter) (Acarina, Tarsonemidae ). Proc. Nat'l. Acad. Sci., 23(2):41-44. 1939. The nuclear cytology of the grass mite, Pediculopsis graminum (Reut.) with special reference to karyomerokinesis. Zeitschrift fiir Zellforschung und Mikroskopische Anatomie, Abt. B., Chromo- soma, Zeitschrift fiir Zellkern- und Chromosomenforschung, 1:51- 103. 1940. Relations of Pediculopsis graminum and Fusarium poae to central bud rot of carnations. Phytopathology, 30( 10) :853-859. CUNLIFFE, F. 1955. A proposed classification of the trombidiform mites (Acarina). Proc. Ent. Soc. Washington, 57:209-218. Desey, M. H. 1849. Ueber mehrere Acarus-Arten. Amtlicher Bericht 25, Versamm- lung der Gesellschaft Deutscher Naturforscher und Aerzte, Aachen, pp. 155-156. 234 THE UNIVERSITY SCIENCE BULLETIN Dest, P. 1838. Sur les mites ou acarus d. vielles criblures. Recueil Agronomique de Tarn et Garonne, 19:149. DEF iwiprt, F. 1861. Lais: nuovo genere di Acari della tribu de Gamasidi. Archivio per la Zoologia, Anatomia, e la Fisiologia, 1:211-215. Evans, G. O. 1952. Terrestrial mites new to Britain. II. Ann. & Mag. Nat. Hist., (12)5(55 ) :660-675. Ewine, H. E. 1912. The origin and significance of parasitism in the Acarina. Trans. Acad. Sci. St. Louis, 21:1-70. 1917. New Acarina. Part IJ—Descriptions of new species and varieties from Iowa, Missouri, Illinois, Indiana, and Ohio. Amer. Mus. Nat. Hist. Bull., 37: 167-169. FRoGGATT, W. W. 1894. Description of a new mite belonging to the genus Heteropus, found in wasps’ nests. Proc. Linn. Soc. New South Wales, (2)9: 259-260. GEBER, E. 1879. Entzundliche Prozesse der Haut durch eine bis Jetzt nicht best- immte Milbe verursacht. Wiener Medizinische Presse, Jahrg. 20 (3 Fortsetzungen ): 1362-1365, 1395-1397, 1428-1429. GISLEN, T. 1940. The number of animal species in Sweden with remarks on some rules of distribution, especially of the microfauna. Lunds Univer- sitets Arsskrift, N. F. Avd. (2)36(2):3-23 (Kungl. Fysiografiska Sallskapets Handlingar N. F. 51(2). Goxsvu, K., G. W. Wuarton, and C. E. YUNKER 1960. Variation in populations of laboratory-reared Trombicula (Lep- totrombidium) akamushi (Acarina: Trombiculidae). Acarologia, 2:199-209. HALLER, G. 1881. Die Mundtheile und Systematische Stellung der Milben. Zoolo- gischer Anzeiger, 4:380-386. 1882. Beitrag zur Kenntniss der Milbenfauna Wiirttembergs. Jahre- shefte des Vereins fiir vaterlandische Naturkunde in Waurttem- berg, 1882, pp. 293-325. Hartic, T. 1834. In G. L. & T. Hartig’s Forstliches und Forstnaturwissenschaftliche Conversationslexikon, Berlin, pp. 733-737. HEALD, F. D. 1906a. Report on the plant diseases prevalent in Nebraska during the season of 1905. Nebraska Exp. Sta. Rep., 19:55, 79. 1906b. Bud-rot of carnations. Science, (N. S.) 23:620. THE FaMILy PYEMOTIDAE 235 Herrs, A. 1926. Okologische Untersuchungen an Pediculoides ventricosus (Newp.) Berlese. Zoologia. Original-Abhandlungen aus dem Gesamtgebiete der Zoologie, 74:3-68. Hirst, S. 1921. On some new or little-known acari, mostly parasitic in habit. Proc. Zool. Soc., London, 1921, (II) :357-378. HorrMann, W. H. 1923a. Uber Triatoma flavida, eine kubanische Reduviide. Munchen medizinische Wochenschrift, 70:603. Also see Zoologischer Bericht, Bd. 3. 1923b. Observations on the occurrence and biology of Triatoma flavida in Cuba (Heteropt., Reduviidae). Ent. News 34:111-113. Hucues, A. M. 1948. The mites associated with stored food products. Ministry of Agri- culture & Fisheries. His Majesty’s Stationery Office, London, pp. 1-168. Hutt, J. E. 1921. Some new mites. The Vasculum, 1921, pp. 17-18. Hunter, W. D. and W. E. Hinps 1905. The Mexican cotton boll weevil. U.S.D.A., Bureau of Ento- mology, Bull. no. 51, 181 pp. Jacor, A. P. 1936a. An undescribed pediculoidid mite from the southern Appalachians. Canadian Ent., 68:82-85. 1936b. Three undescribed pediculoidid mites from the southern Ap- palachians. Canadian Ent., 68:129-133. KararFiaT, H. 1959. Systematik und Okologie der Scutacariden. Beitrige zur Syste- matik und Okologie mitteleuropaischer Acarina, Bd.I: Tyro- glyphidae und Tarsonemini, Teil 2, pp. 627-712. KAUFMANN, O. 1924. Die Weissahrigkeit der Wiesengraser und ihre Bekampfung. Nach- richtenblatt fiir die Deutschen Pflanzenschutzdienst (= Berlin Bio- logische Reichanstalt fiir Land- und Forstwirtschaft), 4(1):1-2. KIRCHNER 1864. Die Milben (Acari) Boéhmens (3 Fortestzung). Lotos, Zeitschrift fiir Naturwissenschaften, 14(8) : 125-126. KoLuer, G. 1882a. Eine Getreide-Milbe als Krankheitserregerin. Pester medizinisch- chirurgische Presse. Also see Biologisches Zentralblatt, III: 127. 1882b. Gabonan élosk6do6. atkafaj Alezai Altal okozott bérbetegség tjabb esetei. Orvosi Hetilap, 26:821-822. Kramer, P. 1877a. Grundzuge zur Systematik der Milben. Archiv fiir Naturgesch- ichte, 43:215-247. 1877b. Zwei Parasitische Milben des Maulwurfs. Ibid., 43:249-295. 236 THe UNIveRSITY SCIENCE BULLETIN KRANtTz, G. W. 1957. Dolichocybe keiferi, a new genus and new species of pyemotid mite, with a description of a new species of Siteroptes (Acarina: Pyemotidae). Ann. Ent. Soc. Amer., 50:259-264. 1959. A redescription of the male of Siteroptes reniformis Krantz 1957 with notes on the family Pyemotidae (Acarina). Ibid., 52:335-337. Krezau, H. 1958. Ejinige neue, nordschwedische Milben-Arten der Gattung Pygme- phorus. Statens Vaxtskyddsanstalt. Meddelanden, 11(72):61-80. 1959a. Pyemotes boylei, eine neue Pymotide aus Hawaii. Zoologischer Anzeiger, 163:148-152. 1959b. Systematik und Okologie der Pyemotiden. Beitrage zur Syste- matik und Okologie mitteleuropaischer Acarina, Band I: Tyro- glyphidae und Tarsonemini, Teil 2, pp. 385-625. LABOULBENE, A. and P. MEGNIN 1885a. Note sur un Acarien utile, le Sphaerogyna- ventricosa Newport. Comptes Rendus Hebdomaires des Séances et Mémoires de la Société de Biologie, (8 )2:282-285. 1885b. Memoire sur le Sphaerogyna ventricosa (Newport). Journal de L’Anatomie et de la Physiologie Normales et Pathologiques de V’Homme et de Animaux, 21:1-18. LAGrEZzE-Fossor, A. and R. J. MONTAGNE 1851. Sur la mite du blé. Recueil Agronomique de la Société des Sciences, Agriculture et Belles Lettres, 32(2). LAWRENCE, R. F. 1940. Three new parasitic mites (Acarina) from South Africa. Jour. of the Ent. Soc. South Africa, 3: 107-111. LICHTENSTEIN, J. 1868. In séance, 9 September, Annales de la Société Entomologique de France, Bulletin des Séances et Bulletin Bibliographique, (4)37: LXXXVI. 1869. In séance, 22 December, Ibid., (4)9: LXXVIII. 1875. In séance, 23 June. Ibid., (5)5:CXXVIII (not CXXXVIII). LinprotuH, C. H. 1957. The faunal connections between Europe and North America. 344 pp. Wiley & Sons, New York, and Almqvist and Wiksell, Stock- holm. Lipeovsky, L. J. 1953. Polyvinyl alcohol with lacto-phenol; a mounting and clearing medium for chigger mites. Ent. News, 64:42-44. LOMBARDINI, G. 1949. Acari Nuovi. Redia, 34:67-74. 1953. Acari Nuovi XXIX. Ibid., 38:181-194. MARLATT, C. L. 1898. The periodical cicada. U.S.D.A. Div. Ent., Bull. No. 14, new series, 118 pp. THE Famity PYEMOTIDAE 237 Mastin, P. 1952. Morphological criteria of phyletic relationships. Systematic Zool., 1:49-70. Mayr, E. 1942. Systematics and the origin of species from the viewpoint of a zoologist. xiv + 834 pp., Columbia University Press, New York. Mayr, E., E. G. Linsuey, and R. L. Usincer 1953. Methods and principles of systematic zoology. ix + 336 pp., McGraw-Hill, Inc., New York, Toronto, London. MICHAEL, A. 1880. On two species of Acarina. Jour, Quekett Microscopical Club, 6(44):113-120. 1890. Report on diseased sugar-cane from Barbados, forwarded by Mr. John R. Bovell. Mites on sugar-cane. Bull., Royal Garden, Kew, no. 40, April, pp. 85-86. Also, see Barbadoes sugar cane mites. Insect Life, 3(1):81. MICHENER, C. D. 1944, Comparative external morphology, phylogeny, and a classification of the bees (Hymenoptera). Bull. Amer. Mus. Nat. Hist. 82(6): 138-326. 1957. Some bases for higher categories in classification. Syst. Zool., 6(4):130-162. MICHENER, C. D., and R. R. Soka 1957. A quantitative approach to a problem in classification. Evolution, 11(2):180-162. MitTcuHELL, R. D. 1955. Anatomy, life history, and evolution of the mites parasitizing fresh-water mussels. Misc. Pubs., Mus, Zool., Univ. of Michigan, 89:1-26. 1958. The evolution of exoskeletons and mite phylogeny. Proc. X Int. Cong. Ent., 1:853-855. Morz, E., and O. MoRGENTHALER 1912. Die Sporotrichum Knospenfaule, eine fiir Deutschland neue Nelken- krankheit. (Zugleich ein Fall von Symbiose). Bericht der Deut- schen Botanischen Gesellschaft, 30:654-662. MonTEROosSO, B. 1934. Osservazioni e Riccerche Sperimentali sulla Biologia di Pedicu- loides ventricosus (Newp.) Berlese. Rivista di Biologia, 16:80-127. MORGENTHALER, O. 1922. Chapter “Bienen und Milben.” Archiv fiir Bienenkunde. 4(2): 45-52. MULLER, J. 1905. Pediculoides avenae n. sp. noch eine Milbenkrankheit des Hafers. Zeitschrift fiir Pflanzenkrankheiten, 15:23-29. 238 THE UNIVERSITY SCIENCE BULLETIN Newport, G. 1850. 1853. NEWSTEAD, 1918. Further observations on the habits of Monodontomerus, with some account of a new Acarus, Heteropus ventricosus, a parasite in the nests of Anthophora retusa. Proc. Linn. Soc. London, 2:70. Paper read March 5, 1850. Also see Annals & Magazine of Natural History, (2)6:395-396. Heteropus ventricosus, n. gen., n. sp. Further observations on the habits of Monodontomerus; with account of a new Acarus ( Hetero- pus ventricosus), a parasite in the nests of Anthophora retusa. Trans. Linn. Soc. London, 21 (2) :95-102. R. and H. M. Duvatu Bionomic, morphological, and economic report on the acarids of stored grain and flour. Royal Soc. Rep. of Grain Pests (War) Committee, 2:19-25. OUDEMANS, A. C. 1906. 1912: 1913. 1915. LO: 1931. 1936. 1937a. Acarologische Aanteekeningen XXI. Entomologische Berichten uitgegeven door de Nederlandsche Entomologische Vereeniging, 27:37-46. Acarologische Aanteekeningen XL. Ibid., 3:231-236. Acarologisches aus Maulwurfsnestern (2 Fortsetzung). Archiv fiir Naturgeschichte, Abt. A, 79(9):68-136. Notizen tuber Acari XXII Reihe (Uropodidae, Tarsonemidae, Bdel- lidae). Ibid., Abt. A, 81(5):38-41. Acarologische Aanteekeningen LXI. Entomologische Berichten uitgegeven door de Nederlandsche Entomologische Vereeniging, 4(93):316 and 331-332. Acarologische Aanteekeningen CXI (Revisie der Tarsonemini Canestrini & Fanzago, 1877). Ibid., 8( 182) :312-331. Neues iiber Pediculoides Targ.-Tozz. 1878. Festschrift fiir Embrik Strand, 1:391-404. Tijdschrift voor Entomologie uitgegeven door de Nederlandsche Entomologische Vereeniging, 80, Verslag, IV-VIII. 1937b. Kritisch historisch Overzicht der Acarologie, Band C: Tarso- Paout, G. 1911. PATAU, K. 1936. Piersic, R. 1901. Rack, G. 1959. nemini, Stomatostigmata, Eleutherengona. I-XXIII and pp. 799- 1348, E. J. Brill, Leiden. Monografia dei Tarsonemidi. Redia, 7:217-281, pl. 7-11. Cytologische Untersuchungen an der haploidparthogenetischen Milbe Pediculoides ventricosus Newp. Berl. Zoologische Jahrbucher, Abteilung fiir Allgemeine Zoologie und Physiologie der Tiere, 56: 277-321. Review of Brucker, 1901. Zoologisches Zentralblatt, 9(1):19-25. Acarophenax dermestidarum sp. n. (Acarina, Pyemotidae), ein Eiparasit von Dermestes-arten. Zeitschrift fiir Parasitenkunde, 19:411-431. THe Famity PYEMOTIDAE 239 REDIKORTSEV, V. V. 1947. The mite Pediculoides ipidarius sp. nov., a parasite of bark beetles. Entomologicheskoe Obozrenie, 129(3-4 ) :247-249. REUTER, E. 1900. Untersuchungen tiber die Ursachen der Weissihrigkeit an Wiesen- grasern in Finland. Acta Societatis pro Fauna et Flora Fennica, 19(1):283-136. 1903. Weissahrigkeit der Getreidearten. Zeitschrift fiir Pflanzenkrank- heiten, 12:324-3388. 1907. Uber die Eibildung bei der Milbe Pediculopsis graminum (FE. Reuter) zugleich ein Beitrag zur Frage der Schlechtsbestimmung. Festschrift fiir J. A. Palmen, no. 7, pp. 3-39. 1909a. Zur Morphologie und Ontogenie der Acariden mit besonderer Beriicksichtigung von Pediculopsis graminum (E. Reuter). Acta Societa Scientiarum Fennicae, 36(4) : 1-287. 1909b. Merokinesis, ein neuer Kernteilungsmodus. Acta Societa pro Fauna et Flora Fennica, 37. Ricuarps, A. G. 1951. The integument of arthropods. ix + 411 pp., Univ. of Minnesota Press, Minneapolis. RovELLI, G. and B. Grassi 1888. Di un singolare acaride, Podapolipus reconditus, nobis. Bolletino della Societa Entomologica Italiana, 20:59-63. Sasa, M. 1947. Contribution to the knowledge of human acariasis. Nisshin Igaku, 34:167-170. SCHAMBERG, J. F. 1910. Grain itch (acarodermatitis urticarioides): a study of a new disease in this country. Jour. Cutaneous Diseases, Feb., 1910, pp. 67-89. SELLNICK, M. 1940. Die Milbenfauna Islands. Goteborgs Kungl. vetenskapsoch vitter- hetssamhilles handligar (B) 6(11):120-125. 1943. Einige Neue Milben aus der Romagna. Bolletino della Societi Entomologica Italiana. Firenze, Genova, 75:22-26. Simpson, G. G. 1961. Principles of animal taxonomy. Columbia Univ. Press, xii + 247 pp. Sinwa, R. N. 1958. A subgeneric revision of the genus Osmia in the western hemi- sphere (Hymenoptera: Megachilidae). Univ. Kansas Sci. Bull. 39(5):211-261. Stewart, F. C. and H. E. Hopcxiss 1908. The Sporotrichum bud-rot of carnations and the silver top of June grass. New York Agr. Exp. Sta. Tech. Bull. no. 7, pp. 83-119. TARGIONI-TOZZETTI, A. 1878. Relazione intorno ai lavori della Stazione di Entomologia Agraria di Firenze per l’anno 1876. Annali dell’Agricoltura, I:244-275. 240 Tuor, Sic. 1934. THe UNIVERSITY SCIENCE BULLETIN Neue Beitrage zur Kenntnis der Invertebraten Fauna von Svalbard. Zoologischer Anzeiger, 107(5-6):114-139. TRAGARDH, I. 1904. Acariden aus Agypten und dem Sudan (Zugleich ein Beitrag zur Kenntnis der Entwicklungsgeschichte der Gattungen Phytoptipal- pus, Pimeliaphilus, Pterygosoma, und Podapolipus). Results of the Swedish Zoological Expedition to Egypt and the White Nile, 1901. Jagerskiold Expedition 20, pp. 1-124. TROUESSART, E. L. 1891 and 1896. Considerations générales sur la classification des acariens TrRYBOM, R. 1893. suivies d'un essai de classification nouvelle. Revue des Sciences naturelles de POuest, 1891, no. 4:289-308. Same title (Suite), Ibid., 1896: 20-54. Physacarus ventricosus Newport, funnen under egendomliga for- hallanden. Entomologisk Tidskrift, 14:121-126. VirzTHuM, G. H. 1914. OAS: 1943. Beschreibung einiger neuen Milben. Zoologischer Anzeiger, 44(7): 315-328. Acarologische Beobachtungen 3 Reihe. Archiv fiir Naturgesch- ichte, Abt. A., Bd. 85, Heft 5:2. | Acarologische Beobachtungen 7 Reihe. Ibid., Abt. A., 89(2): 97-181. Acarologische Beobachtungen 8 Reihe. Archiv fir Naturgeschichte, 90( 10) :1-86. Die Acarofauna der Harzfliisse (Acarologische Beobachtungen 12 Reihe). Sitzungsberichte der Gesellschaft | Naturforschender Freunde zu Berlin, 1927:89-110. Acari. Die Tierwelt Mitteleuropas, 3(7):1-112. Resultats Scientifique de Voyage aux Indes Orientales Néerland- aises. Acarina. _Memoires du Musée Royal D’Historie Naturelle de Belgique, vol. 3, fasc. 5, pp. 5-55. In Bronns Klassen und Ordnungen des Tierreichs, Bd. 5 (Arthro- poda), Abt. IV (Arachnoidea), Buch 5 (Acarina). XI-+ 480 pp. Akademische Verlagsgesellschaft, Leipzig. VON HEYDEN, L. 1835. Etwas uber Xenos. Amtlicher Bericht 12, Versammlung der Natur- forscher und Aerzte, Stuttgart, 1834 (publ. 1835), p. 101. WEBSTER, F’. M. 1883. Observations on the Angoumois grain moth and its parasites. 12th Rep., St. Entomologist on the Noxious and Beneficial Insects of the State of Illinois. Springfield, 1883, pp. 144-154. 1910. A predaceous and supposedly beneficial mite, Pediculoides becomes noxious to man. Ann, Ent. Soc. Amer., 3:15-39. Wess, H. B. 1915. The occurrence of Pediculopsis graminum and the carnation bud- rot in New Jersey. Ent. News, 26:183. THe Famity PYEMOTIDAE 241 Wuarton, G. W. and H. S. FuLLER 1952. WILLMANN, 1942. 1943. 1949. 1952a. 1952b. 1953. 1956. A manual of the chiggers. Memoir Ent. Soc. Washington, no. 4 185 pp. Ge Acari aus Nordwestdeutschen Mooren. Abhandlungen hrsg. Natur- wissenschaftlichen Verein zu Bremen, 32( 1) :163-183. Terrestrische Milben aus Schwedisch-Lappland. Archiv fiir Hy- drobiologie, 40:216-218. Uber eine eigenartige Milben fauna an faulenden Kartoffeln. Ver6ffentlichungen aus dem Museum Naturvolker und Handel- skunde in Bremen, no. 1A, pp. 136-142. Untersuchungen tber die terrestrische Milbenfauna in pannon- ischen Klimagebiet Osterreichs. Sitzungsberichten Osterreichischen Akademie der Wissenschaften, Mathematisch-naturwissenschaftlich Klasse. Abt. I, 160( 1&2) :91-176. Die Milbenfauna der Nordseeinsel Wangerooge. Veroftentlich- ungen der Institut fiir Meeresforschung, 1:139-186. Parasitische Milben an Kleinsiugern. Zeitschrift fiir Parasiten- kunde, 15:392-428. Tarsal Sinnesorgane bei der Gattung Rhagidia und anderen pro- stigmatischen Milben. Zoologischer Anzeiger, 150:215-223. Milben aus dem Naturschutzgebiet auf dem Spieglitzer (Glatzer ) Schneeberg. Ceskoslovenska parasitologie, 3:211-273. r) Wo corr, R. H. 1908. A mite accompanying the bud rot of carnations, Univ. Nebraska Exp. Sta. Bull. no. 103, pp. 25-31. bo — bo THe UNIVERSITY SCIENCE BULLETIN PALPAL SETA 2 —-EXTERNAL DORSAL = i SETA / INTERNAL DORSAL | SETA CAPSULE __ _ __/_ | -GNATHOSOMAL VENTRAL SETA CONDYLAR PROCESS --- Fics. 1-2. Gnathosoma of female Neopygmephorus sp., dorsal (fig. 1) and ventral (fig. 2) aspects. Labelled structures agree with the terminology used in the text. THe Famity PYEMOTIDAE 243 APODEM ANTERIOR MEDIAN APODEME. )CIRCUMGNATHOSOMAL FORAMEN ,APODEMES I _ ,APODEMES IT La Ye {7 ‘#4 47 _2.-- SECONDARY TRANSVERSE A (eZ APODEME NW 4 - POSTERIOR MARGINAL APODEME —APODEMES II ee LAXILLARY SETA = EXTERNAL VENTRAL SETA I~ SS 7 VENTRITE I~ — —___ ; INTERNAL PRESTERNAL SETA~_ EXTERNAL PRESTERNAL SETA~ PI ~~ >-APODEME “|S garooenes w X ERIOR VENTRAL V — ‘POST SN PLATE SN — ae NON SS, 4 \. ANTERIOR AXILLARY SETA IL INTERNAL POSTSTERNAL SETA~ ~~~ 257% Ae EXTERNAL et Se se POSTSTERNAL ~~GENITAL SCLERITE SETA-~~ \ WA 4/ é / —---OPISTHOSOMAL VENTER ; LI | “y;\\ j oo i 7 F oe a fe ) bo Fics. 35-40. THE UNIVERSITY SCIENCE BULLETIN Trochometridium tribulatum, allotype male. Figs. 37-40, right tibiae and tarsi of legs I-IV, respectively, ventral aspect. THE Famity PYEMOTIDAE 263 264 THe UNiversiry SctENCE BULLETIN Fics. 41-42. Caraboacarus sp. female, Lawrence, Kansas, from Agonoderus sp. Fics. 43-44. Acarophenax tribolii, female. India, jute bagging (taken at Puerto Rico in quarantine ). THe Famity Pyemortipar 265 DORSAL Fics, 45-47. Paracarophenax dybasi, female. Composite, holotype and paratypes 1-3. Fig. 47, left tibiotarsus I, ventral aspect. Fics. 48-50. Adactylidium beeri, holotype female. Fig. 50, right tibio- tarsus I, ventral aspect. 266 Tue UNIVERSITY SCIENCE BULLETIN Fics. 51-53. Siteroptes (Siteroptoides) absidatus, holotype female. Fig. 53, right tibiotarsus I, ventral aspect. Fics. 54-56. Pediculaster (Acarothorectes) curculionium, holotype female. Fig. 56, right tibiotarsus I, ventral aspect. THE Famity PyEMOTIDAE 267 Fics, 57-58. Pediculaster (Pediculaster) americanus, female. Composite, lectotype and paratype no. 1. Fics. 59-60. Pygmephorus sp. nr. spinosus, female, Posey Co., Indiana, from nest of Blarina brevicauda. 268 Tue UNIVERSITY SCIENCE BULLETIN | \ ti ff elk Wg A bs \ Hy/) / ) C325 Rae ek ee os Pe eee 320 Nomia/m: melanderi: = oscar eee 323 Nomia im. howardts.@ 274... So Pk eee 332 Nomia-robinsoni on oS ee ee I ee eee 334 BIBLIOGRAPHY 2202255020) Geese Da Le is nS 338 INTRODUCTION This paper is a taxonomic revision of the American subgenus Acunomia and the American species of Curvinomia of the genus Nomia. Species of these two subgenera can be differentiated from any other North American bees by their brilliant, metallic colored abdominal bands. Other subgenera may have abdominal bands of pale pubescence but do not have integumental bands. The sub- genera of Nomia are in a state of taxonomic confusion, and it seems best not to make any subgeneric changes until the large African fauna has been more fully studied (Michener, 1961). When this has been done it is likely that the subgenera Acunomia, Curvi- nomia and Hoplonomia will be included in Nomia s. str. All of them have colored integumental abdominal bands. I follow Cross (1958a) in the placement of the species in Acunomia and Para- nomina (= Curvinomia). There have been nineteen specific and four subspecific names proposed for Acunomia and the North American species of Curvi- nomia. When this study was begun, fourteen specific and three subspecific names were being used. I consider nine species and three subspecies as valid. Seven names have been placed in synon- ymy for the first time, and one species is described as new. Three other names appear in new subgeneric combinations, one of these being elevated from previous synonymy. A total of 4,781 bees were examined during this study. Acunomia seems very closely related to Curvinomia and probably arose from it since our Curvinomia became separated from the Palearctic fauna (Cross, 1958a). Since Curvinomia reaches only slightly above 44° north latitude in North America, our species must have been separated from the Palearctic ones at least since the Pliocene, for that was the last time portions of Alaska had a tem- perate climate (Hibbard, 1960). BANDED SUBGENERA OF NOMIA IN AMERICA 279 ACKNOWLEDGMENTS I would like to express my deep appreciation to Dr. C. D. Michener for his interest, encouragement and guidance during the course of this study, for examining the types in the U. S. National Museum, and for reading and making suggestions in the manu- script. Appreciation is also due Dr. G. W. Byers, who has given much helpful advice and has read and commented on the manu- script. A debt of gratitude is due Dr. J. G. Rozen, Jr., of the American Museum of Natural History, Dr. P. H. Timberlake of the Citrus Experiment Station, Riverside, California, Dr. K. V. Krom- bein of the U. S. Department of Agriculture, and Mr. H. B. Leech of the California Academy of Sciences, San Francisco, who have allowed the types in their care to be examined. Appreciation is due Dr. M. W. R. deV. Graham of Oxford University for comparing the type of N. cressoni with specimens sent to him. I wish to express thanks to the following persons who have rendered services or loaned material in their care: W. T. Atyeo, University of Ne- braska; W. F. Barr, University of Idaho; G. E. Bohart, U. S. Legume Seed Research Laboratory, Logan, Utah; H. R. Burke, Texas Agri- cultural Experiment Station, College Station; G. D. Butler, Univer- sity of Arizona; L. Chandler, Purdue University; E. F. Cook, Uni- versity of Minnesota; E. A. Cross, Northwestern State College, Natchitoches, Louisiana; W. A. Drew, Oklahoma State University; J. G. Edwards, San Jose State College; W. R. Enns, University of Missouri; R. L. Fischer, Michigan State University; R. C. Funk, University of Kansas; L. kK. Gloyd, [linois Natural History Survey; C. E. Hopla, University of Oklahoma; P. E. Hunter, University of Georgia; P. D. Hurd, Jr., University of California, Berkeley; M. T. James, Washington State University; G. F. Knowlton, Utah State University; J. N. Knull, Ohio State University; H. C. Knutson, Kansas State University; K. V. Krombein, U. S. Department of Agriculture; L. C. Kuitert, Agricultural Experiment Stations, Gaines- ville; J. L. Laffoon, Iowa State University; R. J. Lavigne, University of Wyoming; K. W. MacArthur, Milwaukee Public Museum; A. T. McClay, University of California, Davis; H. E. Milliron, Canada Department of Agriculture, Ottawa; T. B. Mitchell, North Carolina State College; R. A. Morse, Cornell University; R. L. Post, North Dakota State University; J. A. G. Rehn, Academy of Natural Sci- ences of Philadelphia; H. G. Rodeck, University of Colorado Mu- seum; J. G. Rozen, Jr., American Museum of Natural History; H. C. Severin, South Dakota State College; G. I. Stage, California 280 THE UNIVERSITY SCIENCE BULLETIN Academy of Sciences, San Francisco; W. P. Stephen, Oregon State University; P. H. Timberlake, Citrus Experiment Station, Riverside, California; P. F. Torchio, Oregon State University; F. S. Truxal, Los Angeles County Museum; G. E. Wallace, Carnegie Museum; H. V. Weems, Jr., State Department of Agriculture, Gainsville, Florida; F. G. Werner, University of Arizona; S. L. Wood, Brigham Young University; and I. H. H. Yarrow, British Museum (Natural History). Appreciation is also due all those who collected the material studied, although space does not permit thanking them individually. METHODS The synonymy is given in abbreviated form, since the complete reference can be found in the bibliography. All papers known to refer to the species or subspecies in question are listed, whether or not they are strictly taxonomic in nature. After each citation is an abbreviation or word in parentheses that indicates the main emphasis of the paper as it pertains to the particular species or subspecies. “Taxonomic” is abbreviated “tax.”; if followed by @ there is a description of the female, and if by ¢, a description of the male is given. The other abbreviations used are as follows: alf. = alfalfa pollination, biol. = biology, fls. = flower record, para. — parasites and predators, and rec. = locality record. The method of Moure and Michener (1955) in numbering char- acters in the descriptions is used for easy comparison. Specific and subspecific descriptions use the same numbers but separate numbering systems are used for the two subgenera. The wings are measured from the tip of the marginal cell to the wing base along the anterior margin. Tergum 1 refers to the first metasomal tergum, and the other terga and sterna are numbered accordingly. This system does not correspond to the strictly morphological sys- tem, in which the first metasomal tergum would be abdomina! tergum 2, but it is generally used (Moure and Michener, 1955) and presents less opportunity for confusion. In using the terms gonocoxite, gonostylus, etc., I do not wish to imply the appendicular origin of these structures, but am merely following with established practice. The terms used in the de- scriptions of sterum 5 and the genitalia are shown in figures 24 and 30. When referring to Cockerell’s types, currently accepted terms for the kinds of types will be used. Cockerell used type when referring to a holotype and his cotype is the same as paratype. BANDED SUBGENERA OF NOMIA IN AMERICA 281 In the lists of localities, an asterisk refers to a record in the litera- ture. A question mark indicates a locality that was not found on any map examined. After the locality the number of each sex examined from that locality is given. If no number precedes the sex symbol, there is but one specimen of that sex. An attempt was made to use standardized, easily found place names. When needed, more specific information is given in parentheses for certain localities. In the lists of flower records, those taken from the literature are followed by an asterisk. After the name of each flower is given the number of specimens of each sex taken. If no number precedes the sex symbol there is but one specimen of that sex. An attempt was made to use currently accepted plant names. TAXONOMIC CHARACTERS In the descriptions all variations found have been described, even though not found to be taxonomically significant. In the males the hind tibiae, genitalia and the fifth modified sterna offer the best specific characters. In the females the punctation of the metasomal terga provide the best specific characters. The integu- mental color was found to be highly variable, ranging from black to red-brown in all species. The color varies in a haphazard fashion. Aging in museums may have some effect on color. In general the very old specimens were the reddest, while those col- lected recently were the blackest, although there was a little varia- tion in specimens of the same age. It is not known whether it is light, paradichlorobenzene or some other factor that causes museum specimens to lose their black color. The metepisternum is highly variable in wrinkle size and pattern. The anterior groove of the propodeum is highly variable in shape and wrinkle pattern. The color of the abdominal bands varies greatly; therefore, this char- actor should be used with caution. These bands tend to lose their color permanently if they become wet, as in a relaxing or killing jar. Band width is a more reliable character. The amount of black hair on the metasoma is highly variable. On older specimens the hair is often worn away, especially on the clypeus and metanotum, and these specimens may not fit the descriptions in this respect. In describing the punctation of the metasomal terga it is the central portion of the tergum that is referred to, unless otherwise stated. 982 Tue UNIVERSITY SCIENCE BULLETIN Kry TO THE SPECIES AND SUBSPECIES OF ACunomia AND NortH AMERICAN Curvinomia 1. Males: Antenna thirteen-segmented; hind tibia and femur enlarged; fifth metasomal sternum with two large hairy pads................... Females: Antenna twelve-segmented; hind tibia and femur unmodified; fifth metasomal sternum without hairy pads....................:... . Hind tibia produced anteriorly and distally into large lobe hiding all or almost all of tibial spurs in side view (fig. 7); antenna tapering and if bent backward, reaching beyond thorax (subgenus Acunomia)......... Hind tibia triangular, not hiding most of tibial spurs in side view (fig. 1); antenna not tapering and if bent backward not reaching beyond thorax (Subgenus Curvinomia)) aus ese ie Mee ae) soa 3. Large, front wing length (see section on methods) 10.0 mm. or more; mandible with large rounded lobe on posterior margin (species norton2) . . Medium-sized, front wing length less than 10.0 mm.; mandible with toothlike lobe or lobe absent on posterior margin. .................-. 4. Hind tibial lobe with raised area of outer surface limited to basal half, raised area not forming distinct ridge; anterior margin of general area, DONG ICG; DATE size) sce oes acco eS OA eee n. nortont Hind tibial lobe with raised area on outer surface extending toward tip, raised area forming distinct ridge: anterior margin of general area, along eye, with short, appressed, plumose hairs.................. n. cressoni 5. Hind tibial lobe with apex rounded (fig. 9); pronotum without posterior, dorsal carina; (species melandert) 7. .- 9.2. dos nein cee ae eee Hind tibial lobe with deep apical emargination (fig. 11); pronotum with strong posterior, dorsal carina...................0--ee ee eeee robinson 6. Colored band on tergum 1 broken or much reduced toward center. m. melandera Colored band on tergum 1 present and not reduced toward center. m. howardi 4..Clypealimarginswathimedian tootht= Ge.) she ose ee angustitibialis Clypeal margin -withoutamedian tooth: 2. . 5.00.2. a. oe 8. Punctures on tergum 4 close, minute, much smaller than those on ter- UID SD ee Ee Ses ee A tN ae aa Punctures on tergum 4 not close, large, about the same size as those on bo ero Qe as Se i ee a aa a aA a 9. Hind femur three times as long as broad...................... maneet Hind femur twicesas long as broads... ..)4..4 5. ee 10. Posterior tibia with apex yellow (fig. 5)......................2.0. fox Posterior tibia with only tip of lobe yellow (fig. 2)........... fedorensis 11. Transverse groove on tergum 2 shallow and U-shaped in cross-section. . Transverse groove on tergum 2 deep with anterior face nearly vertical, posterior face sloping gently upward.................... t. tetrazonata 12. Apical, depressed area on tergum 2 about as wide as rest of postgradular BUR cY Maman (Aran Uaenie Ale Ace a Ee Ge Mca Wai lln iN Na Ue main nea Asai in a lola @ t. wvaldensis Apical, depressed area on tergum 2 less than one-half as wide as rest of ? . . . posteraduilanianea pe 2h C ore GRO es Pere at Se een uniwersitatis 13. Middle tibial spur finely and evenly serrate along entire length (subgenus CUrvrmomeg) (2 55 20 A OE SBS NN is Ste CLD Bers cae Middle tibial spur with few, uneven, coarse teeth along apical third (sub- genus Alewiorniea) ise eee ee tins ALE Si at usheeuncc emule nani Cel en 14. Tergum 1 with apical colored band (species melandert)............... Tergum 1 without apical colored band..................-...0-0000- 13 11 10 BANDED SUBGENERA OF NOMIA IN AMERICA 283 15. Colored band on tergum 1 one-third width of band on tergum 2. m. melandert Colored band on tergum 1 one-half width of band on tergum 2. m. howardi 16. Large, front wing length (see section on methods) more than 8.0 mm.; pronotum without posterior, dorsal carina (species nortont)........... 17 Medium-sized, front wing length less than 8.0 mm.; pronotum with pos- COTO AC ONS ARCATA eves seo ines Or eae eyeaedes snare aoa olay mae sero fom. outscer robinsont 17. Anterior margin of genal area, along eye, bare.............. n. nortont Anterior margin of genal area, along eye, with short, appressed, plumose INATTS VON LTR Rater My ete PS tenes: MANA NUD n Os Ln Mens: Gal eee Ca Day. n. Cressont 18. Raised areas on terga 1 to 3 nearly impunctate and highly polished. . .fox7z Raised area on terga 1 to 3 usually well punctured and not polished... 19 19. Tergum 3 with punctures close and minute, much smaller than those in CembersOls bere ume saver a Syed Ue A i Rec ee Sal Pein iar oneal 20 Tergum 3 with punctures medium to large, not close and not much smaller than=thosean- center of tercumy2.> Sa poe soa cede eee en ce ie eee oes ee 21 20. Postgradular area depressed below level of pregradular area on tergum 2. maneet Postgradular area not depressed below level of pregradular area on ter- AUT etme ate Me MleveatnaNSERIA UNCER oe (aa a eM eae a fedorensis 21. Centers of terga 1 and 2 with punctures minute, much finer than those @lemnleS oc uC UMN sl rae pees Ee re ree CUA ne Ue oy eer amen t. waldensis Centers of terga 1 and 2 with punctures fine to coarse, not much finer thanuthoseoimesoscutumie ss oace nee cae ee tic ee oer iae olen. 22 22. Colored band on tergum 3 narrow, 0.33 mm. in width................ 23 Colored band on tergum 3 wide, 0.50 mm. or more in width. angustitibialis 23. Wing length about 9 mm.; midwestern United States...... uniersitatis Wing length about 7 mm.; southwestern United States and Mexico. t. tetrazonata Subgenus Curvinomia Michener Nomia (Paranomia) Friese, 1897, p. 48 (not Paranomia Conrad, 1860) (type species: N. chalybeata Smith); Cockerell, 1905, p. 322; Cockerell, 1910b, p. 289; Cockerell, 1930, p. 12; Blair, 1935, pp. 208 and 211; Sandhouse, 1943, p. 584. Paranomia: Ashmead, 1899, p. 88. Nomia (Paranomina) Michener, 1944, p. 251 (not Paranomina Hendel, 1907), (new name for Paranomia Friese); Cross, 1958a, p. 1263. Nomia (Curvinomia) Michener, 1944, p. 251 (type species: N. californiensis Michener, = N. t. tetrazonata Cockerell); Michener, 1951,/p. 1127. Except for the original reference, only papers dealing with North American Curvinomia are cited in the synonymy above. Since Paranomia and Paranomina are preoccupied, this subgenus takes the name Curvinomia, the only other available name. 984 THe UNIVERSITY SCIENCE BULLETIN The characters that separate the North American Curvinomia from Acunomia are given below: Mate. Antenna not tapering and not reaching beyond thorax when bent backward, last flagellar segment blunt or angular. Pos- terior tibia triangular in shape, not hiding most of tibial spurs in side view (fig. 1). Tergum 1 never with colored band. Sternum 3 without ventrolateral process. Sternum 5 with setae in row behind lateroapical hairy pads not completely fused; median apical process arising from median longitudinal ridge only posteriorly (fig. 12). Penis valve without ventral opening (fig. 18); gonostylus without small, inward-pointing, hooklike process at base. Fremate. Middle tibial spur finely and evenly serrate along entire length. Tergum | never with colored band. Nomia (Curvinomia) maneei Cockerell (Figs. 1, 12 and 18; Map 1; Table 5) Nomia maneei Cockerell, 1910a, p. 276 (tax., 2 ); Cockerell 1930, p. 18 (tax.); Cockerell, 1934b, p. 7 (tax.); Brimley, 1938, p. 457 (rec.); Fattig, 1943, p. 6 (para.); Muesebeck et al., 1951, p. 761 (paras): Nomia (Acunomia) maneei: Michener, 1951, p. 1128 (cat); Mitchell, 1960, p. 508 (tax., 6, @ ). Paranonia [sic] maneei: Snyder, 1957, p. 58 (rec. ). Nomia (Paranomina) maneei: Cross, 1958, p. 1266 (tax.). This species can be differentiated from other North American Curvinomia by the minute, close punctures on terga 3 and 4, the narrow, bright green abdominal hands, and in the male by the only slightly modified hind femur and tibia. Mate. Structure: (1) Clypeal margin broadly rounded in center with narrow submarginal, impunctate area; clypeus with large, rounded, longitudinal ridge. (2) Metepisternum with many fine to a few coarse wrinkles. (3) Propodeum with fine punctures on pos- terior face. (4) First m-cu interstitial with to several vein widths basal from first r-m. (5) Posterior femur three times as long as broad. (6) Posterior tibia slightly expanded toward apex (fig. 1). (7) Tergum 1 with medium-sized punctures except for band of minute punctures at base of apical, depressed, hyaline portion; median width of band one-half width of depressed area. (8) Ter- gum 2 with punctures like those of center of tergum 1; groove between postgradular and pregradular areas deep, its anterior face nearly vertical, posterior face sloping gently upward. (9) Terga 3 BANDED SUBGENERA OF NOMIA IN AMERICA 285 and 4 much more finely punctate than center of tergum 1, except raised area of tergum 3 which has punctures like those of tergum 1. (10) Sternum 5 with lateral, longer part of median, apical process about as long as wide and rounded at apex, processes widely sepa- rated; lateroapical hairy pads with hairs sometimes meeting along midline, hairs in center of pad not darkened, pads half round in shape; median, basal emargination much more than twice as wide as deep, V-shaped (fig. 12). (11) Gonocoxite with apex of dorsal recurved part bent outward and rounded. (12) Gonostylus with plate at its base with small bumps on inner side. (13) Exposed part of penis valve two-thirds as wide as long, tip broad and pointed, inner subapical margin with small rounded lobe, extreme lateral part with anteriorly pointing tooth, anteroventral process J-shaped (fig. 18). Pubescence: (14) Terga 4 to 7 with some black hairs, sometimes absent on tergum 4; tergum 3 with hairs all white. (15) Gonocoxite with dorsal recurved part with large, stout hairs at tip arising from well behind inner margin. Color: (16) Mandible translucent orange to red-brown with red tip and black to orange base. (17) Antenna orange to red-brown below, black to red-brown above. (18) Legs brown to red-brown with tarsi and tip of hind tibial lobe which may be orange to yellow. (19) Stigma darkened, darker in color than most wing veins. (20) Colored band of tergum 2 about 0.25 mm. in width; color bright green with pink or purple iridescence. FEMALE. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is broadened distally. (22) Clypeal margin truncate in center. (23) Head and thorax with fine punc- tures; mesoscutum with spaces between punctures minutely rough- ened. (24) Metepisternum with several medium to fewer coarse wrinkles. (25) Tergum 1 with close, fine punctures except for band of minute punctures at base of apical, depressed, hyaline portion; median width of band three-fourths width of depressed area; center of tergum not highly polished. (26) Tergum 2 with punctures same size as those of center of tergum 1 or smaller; tergum 2 not highly polished; postgradular area abruptly depressed below level of pregradular area. (27) Terga 3 and 4 with minute, close punctures, much smaller than those of tergum 2. Pubescence: (28) Body hairs white with some black hairs on terga 2 to 6. Color; (29) Antenna orange to red-brown below, black to red-brown above. (30) Stigma darkened, darker in color than most wing veins. (31) Colored band of tergum 2 about 0.33 mm. in width; color bright green with pink or purple iridescence. 986 THE UNIVERSITY SCIENCE BULLETIN Type Materiau. The female holotype of maneei, Southern Pines, North Carolina, June 19, 1909 (A. H. Manee), could not be located. This specimen was in the Birkman collection at the time of de- scription. I have seen a female in the Illinois State Natural History Survey collection and another female in the Citrus Experiment Station collection that bear these same label data, but neither of these specimens seems to be the holotype of maneei. The Illinois State Natural History Survey collection also has two males with the same label data. Since Cockerell apparently described maneei from a single female these other specimens are only topotypes. The designation of a neotype does not seem warranted. Rance. N. maneei ranges from North Carolina to Florida west- ward nearly to the Mississippi River (map 1). One specimen has been collected in New Jersey. This range does not approach that of any other Curvinomia. List oF Locatities. ALABAMA. Colbert Co.: Sheffield ¢. Fiormpa. Alachua Co.: Gainesville @; Dade Co.: Miami 4; Duval Co.: Jacksonville Beach ¢; Levy Co. 24; Orange Co.: Winter Park ¢@; Pasco Co.: Lacoochee ¢; Polk Co.: Fort Meade 6; Suwanee Co.: Suwanee Springs @ 6. Grorcra. Dade Co.: Head River 3924; Decatur Co.: Bainbridge 6, Spring Creek 24; DeKalb Co.: Stone Mountain ¢; Dougherty Co.: Albany 4; Floyd Co.: Rome ? ¢; Franklin Co.: Royston 2 ¢ ; Stephens Co.: Toccoa @; Warren Co.: Warrenton ¢. Mussissrppi. Harrison Co.: Gulfport ¢°; Lowndes Co.: Columbus @. New Jersey. Burling- ton Co.: Rancocas @. NortTH CaRroLina. Craven Co.: Croatan 24; Harnett Co.: Lillington 956; Lenoir Co.: LaGrange 9°; Macon Co.: Highlands (Horse Cove) ¢; McDowell Co.: Marion 2; Moore Co.: Southern Pines 8 2 18 6; New Hanover Co.: Caro- lina Beach 98, Wilmington 29 ¢, Wrightsville Beach ¢; On- slow Co. 22: Dixon;* Richmond Co.: Hoffman 49; Swain Co.: Bryson City 10928; Wake Co.: Raleigh 29. SourH CaRoLina. Abbeville Co.: Calhoun Falls ¢; Aiken Co.: Aiken ¢. FLower Recorps. ASCLEPIADACEAE: Asclepias tuberosa 92. LEGUMINOSAE: Desmodium ciliare 2 6, Galactia sp. 2? 6, Lespe- deza capitata 8, L. repens 102 6, Phaseolus polystachios 23 ¢, Strophostyles 9. PotyconacearE: Eriogonum baileyi 2. Parasites. Pseudomethoca vanduzeei (Hymenoptera: Mutilli- dae) has been recorded from the pupa of maneei (Fattig, 1943). SEASONAL Activity. Table 5 gives the seasonal activity of the adults. BANDED SUBGENERA OF NOMIA IN AMERICA 287 Nomia (Curvinomia) fedorensis Cockerell (Figs. 2, 13 and 19; Map 2; Table 5) Nomia foxii: Birkman, 1899, p. 244 (rec.) (misidentification ). Nomia fedorensis Cockerell, 1910a, p. 277 (tax., 2, ¢ ); Cockerell, 1930, p. 13 (tax.); Cockerell, 1934b, p. 4 (biol.). Nomia (Acunomia) fedorensis; Michener, 1951, p. 1127 (cat.). Nomia (Paranomina) fedorensis: Cross, 1958a, p. 1266 (tax.). This species can be differentiated from other North American species of Curvinomia by the very small, close punctures on terga 3 and 4, the bright emerald green abdominal bands, and in the male by the greatly expanded hind femur and tibia. Mate. Structure: (1) Clypeal margin broadly rounded in center with narrow submarginal, impunctate area; clypeus with distinct longitudinal carina. (2) Metepisternum with many fine to fewer medium wrinkles. (3) Propodeum with medium-sized punctures on posterior face. (4) First m-cu nearly interstitial with to several vein widths basal from first r-m. (5) Posterior femur twice as long as broad. (6) Posterior tibia greatly expanded toward apex (fig. 2). (7) Tergum 1 with coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band one-third to one-half width of depressed area. (8) Tergum 2 with punctures like those of center of tergum 1; groove between postgradular and pregradular area deep and V-shaped. (9) Terga 3 and 4 much more finely punctate than center of tergum 1, except raised area of tergum 3 which has punctures like those of tergum 1. (10) Sternum 5 with lateral longer part of median, apical process absent, process one-half as long as wide and truncate at apex, processes nearly converging along midline; lateroapical hairy pads with hairs sometimes meet- ing along midline, hairs in center of pad not darkened, pads half round in shape; median, basal emargination from as wide as deep to twice as wide as deep, U-shaped, or tending toward V-shaped (fig. 13). (11) Gonocoxite with apex of recurved part bent out- ward and rounded. (12) Gonostylus with plate at its base papil- lose on inner side. (13) Exposed part of penis valve two-thirds as wide as long, tip broad and pointed, inner subapical margin with small rounded lobe, extreme lateral part with anteriorly pointing tooth, anteroventral process J-shaped (fig. 19). Pubes- cence: (14) Terga 3 to 7 with some black hairs. (15) Gonocoxite with dorsal recurved part with large, stout hairs at tip arising well behind inner margin. Color: (16) Mandible translucent yellow 288 THe UNIVERSITY SCIENCE BULLETIN to red-brown with red tip and brown to yellow base. (17) Antenna yellow to orange below, red-brown to orange above. (18) Legs black to red-brown with tip of hind tibial lobe yellow. (19) Stigma light in color like most wing veins. (20) Colored band of tergum 2 about 0.33 mm. in width; color bright emerald green with pink, purple, gold or yellow iridescence. FEMALE. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is broadened distally. (22) Clypeal margin truncate in center. (23) Head and thorax with fine punc- tures; mesoscutum with spaces between punctures smooth. (24) Metepisternum with many fine to fewer medium-sized wrinkles. (25) Tergum 1 with close, coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band one-half width of depressed area; center of tergum not highly polished. (26) Tergum 2 with punctures same size as those of center of tergum 1 or smaller; tergum 2 not highly polished; postgradular area not depressed below level of pregradu- Jar area. (27) Terga 3 and 4 with minute, close punctures, much smaller than those on tergum 2. Pubescence: (28) Body hairs white or yellowish with some black hairs on terga 3 to 6. Color: (29) Antenna orange below, black to red-brown above. (380) Stigma darkened or light in color like most wing veins. (31) Col- ored band on tergum 2 about 0.33 mm. in width; color bright emerald green with purple, red or pink iridescence. VariATION. Little variation has been noted in the characters studied in the few specimens of this species examined. A female from Vera Cruz, Mexico, has the yellowish hairs mentioned in character 28; other specimens have whitish hairs. Type MaterraL. The female holotype of fedorensis, Fedor, Lee County, Texas, June 7, 1909 (G. Birkman), is in the U. S. National Museum, type number 55864. A male paratype is also there, with the same label data except the date (June, 1910). Rance. N. fedorensis occurs in southern and central Texas with a few specimens known from southwestern Kansas and Vera Cruz, Mexico (map 2). List oF Locauitirs. Kansas. Clark Co.3978. Texas. Atas- cosa Co.: Lytle 946; Bastrop Co.: McDade 24; Bexar Co. 11¢ 4; Brooks Co.: Fulfurrias ¢; Cameron Co.: Boca Chica ¢; Dimmit Co.: Carrizo Springs ?; Eastland Co.: Romney 9442; Goliad’ Co. °: Weser?; Kenedy Co: Sarita (18 miriSajes BANDED SUBGENERA OF NOMIA IN AMERICA 289 (29 mi. S.) 944; Lee Co.: Fedor 4934; San Saba Co.: San Saba Camp ¢. Mexico. Veracruz: Veracruz ?3¢. FLower Recorps. ComposiraE: Gaillardia ¢, Helianthus sp. 6, genus? 2. LecumINosaE: Cassia sp.,* Dalea grisea 246, Mimosa, sp. 22 8, Rhynchosia americana,* Tephrosia virgini- ana ¢. Brotocy. Cockerell (1934b) mentions a nesting site of fedoren- sis. Nothing else is known of the biology of this species. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Curvinomia ) universitatis Cockerell (Figs. 3, 14 and 20; Map 1; Table 5) Nomia punctata var. (?) Fox, 1898, p. 184 (not Smith, 1875) (tax., 2 ); Cockerell, 1930, p. 14 (tax.). Nomia foxii Dalla Torre, 1896, p. 167 (new name for N. punctata Fox) (in part) (cat.); Cockerell, 1906a, p. 298 (in part) (rec.) (misidentifications ). Nomia punctata: Cockerell, 1898d, p. 31 (in part) (tax.); Swenk, 1907, p. 293 (ree. ). Nomia universitatis Cockerell, 1908, p. 334 (tax., ¢ ); Cockerell, 1910a, p. 277 (tax.); Cockerell & Robbins, 1910, p. 194 (tax.); Cockerell, 1925b, p. 233 (rec.); Blair, 1930, p. 17 (tax.); Michener, 1936, p. 4 (tax. ). Nomia (Paranomia) wuniversitatis: Cockerell, 1910b, p. 297 (tax. ); Cockerell, 1930, p. 14 (tax.). Nomia (Acunomia) universitatis: Michener, 1951, p. 1128 (cat.). Nomia universatis: Stevens, 1951, p. 64 (lapsus calami) (tax.). Nomia (Paranomina) universitatis: Cross, 1958a, p. 1266 (tax.). Fox described punctata from three female specimens, one from each of the following localities: Denver, Colorado; Vega San Jose, New Mexico; and Big Stone City, South Dakota. Only the speci- men from New Mexico has been located. This specimen belongs to the species that has been considered mesillensis (= foxii), the least punctate North American banded Nomia. The other speci- mens probably were universitatis, since that is the only Curvinomia known to occur in northern Colorado or South Dakota. The much larger N. nortoni is the only similar bee known from the same area. Fox recognized nortoni and included it in his key (1893). Fox (1893) considered one of the syntypes of punctata a possible “variety because of its strong punctation. Therefore, the specimen 290. THE UNIversiry SCIENCE BULLETIN from New Mexico could not be this “variety.” Since universitatis is strongly punctate it is here considered as Fox’s “variety.” Cock- erell (1908, 1930) came to a similar conclusion. One problem with this reasoning is that Fox’s variety included a single specimen, but on the basis of the localities stated above, both remaining syntypes should be strongly punctate. This matter can be settled only if these two missing specimens are found, which seems unlikely. N. universitatis can be distinguished from other North American Curvinomia by its coarse punctation and narrow abdominal bands, Mate. Structure: (1) Clypeal margin broadly rounded in cen- ter, with wide submarginal, impunctate area; clypeus with weak longitudinal carina. (2) Metepisternum with many fine to a few coarse wrinkles. (3) Propodeum with coarse punctures on posterior face. (4) First m-cu several vein widths basal from first r-m. (5) Posterior femur two and one-half times as long as broad. (6) Posterior tibia greatly expanded toward apex (fig. 3). (7) Tergum 1 with coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band one-third width of depressed area to reduced to a few scattered punctures. (8) Tergum 2 with punctures like those of center of tergum 1; groove between postgradular and pregradular area shallow and U-shaped. (9) Terga 3 and 4 with punctures like those in center of tergum 1. (10) Sternum 5 with lateral, longer part of median, apical process twice as long as wide and truncate to rounded at apex; processes widely separated; lateroapical hairy pads with hairs considerably overlapping along midline, hairs in center of pad not darkened, pads crescent-shaped; median basal emargination about twice as wide as deep, U-shaped tending to- wards V-shaped (fig. 14). (11) Gonocoxite with apex of dorsal recurved part bent outward and rounded. (12) Gonostylus with plate at its base with small bumps on inner side. (13) Exposed part of penis valve two-thirds as wide as long, tip broad and pointed, inner subapical margin rounded, extreme lateral part with anteriorly pointing tooth, anteroventral process J-shaped (fig. 20). Pubescence: (14) Terga 5 to 7 with some black hairs, some- times absent on tergum 5; terga 3 and 4 with hairs all white. (15) Gonocoxite with dorsal recurved part with large, stout hairs at tip arising from well behind inner margin. Color: (16) Mandible red-brown with red tip and black base. (17) Antenna orange to red-brown below, black above. (18) Legs red-brown with tarsi and hind tibial lobe yellow. (19) Stigma darkened or light in BANDED SUBGENERA OF NOMIA IN AMERICA 291 color like most wing veins. (20) Colored band of tergum 2 about 0.25 mm. in width; color light green with purple or pink iridescence. FreMALE. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is broadened distally. (22) Clypeal margin slightly emarginate in center. (23) Head and thorax with coarse punctures; mesoscutum with spaces between punctures smooth or minutely roughened. (24) Metepisternum with many fine to fewer coarse wrinkles. (25) Tergum 1 with close, coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band two-fifths width of depressed area to absent in center and reduced to two rows at sides; center of tergum not highly polished. (26) Tergum 2 with punctures same size as those of center of tergum 1; tergum 2 not highly polished; postgradular area not depressed below level of pregradular area. (27) Terga 3 and 4 with coarse punctures like those of center of tergum 2 except punctures toward base of tergum 4 smaller than on rest of tergum. Pubescence: (28) Body hairs white with some black hairs on terga 4 to 6, black hairs may be absent on tergum 4. Color: (29) Antenna orange to red-brown below, black to red-brown above. (30) Stigma darkened or light in color like most wing veins. (31) Colored band of tergum 2 about 0.33 mm. in width; color yellow or light green with pink, red or purple iridescence. Type MartertaLt. The male holotype of wniversitatis, Boulder, Colorado, July 14, 1908 (Paul M. Dean), is in the U. S. National Museum, type number 55859. A male paratype, Boulder, Colorado, July 6, 1908 (T. D. A. Cockerell), was not located. Rance. N. universitatis ranges from North Dakota to Texas be- tween the Rocky Mountains and the Mississippi River (map 1). List oF Locatities. CoLorapo. Baca Co.: Two Buttes 2 ¢ ; Bent Co.: Prowers (N. of ) 2?; Boulder Co. 39: Boulder 39°32, (2 mi. N.) ¢, (Boulder Canyon) 2?; Denver Co.: Denver 2 ¢; FE] Paso Co.: Colorado Springs 6; Huerfano Co.: La Veta 9¢; Jefferson Co.: Golden 24; Larimer Co.: Fort Collins 3944; Logan Co.: Peetz 28; Iowa. Lyon Co.: Little Rock 2¢. Kan- sas. Bourbon Co.: Fort Scott 39346; Clark Co.: State Park ¢; Douglas Co.: Baldwin City ¢, Lawrence ¢; Ellis Co. 2?; Geary Co. 2°; Graham Co. 2; Gray Co. 2; Jewell Co. 6; Marshall Co. 29: Marysville 2 6; Morton Co. ?; Riley Co. 964: Manhattan ¢; Scott Co.: Scott City ¢; Sheridan Co. ? ¢; Stanton Co. ?. NeE- 292, THe UNIVERSITY SCIENCE BULLETIN BRASKA. Adams Co.: Hastings @ ¢; Cherry Co.: Valentine (Fort Niobrara) ¢; Cuming Co.: West Point ¢; Custer Co.: Broken Bow @; Dixon Co.: Maskell ¢; Douglas Co.: Omaha @? 4; Dundy Co. ¢@; Fillmore Co.: Fairmont 4¢; Lancaster Co.: Lincoln 7924; Scotts Bluff Co.: Mitchell @ ¢. Nrw Mexico. Bernalillo Co.: Sandia Mountains ?. Norra Dakota. Slope Co.: Mar- marth.* Sourn Daxora. Brookings Co.: Brookings ¢; Custer Co.: Pringle ¢; Davison Co.: Mitchell ?; Grant Co.: Big Stone City;* Fall River Co.: Hot Springs 36. Trxas. Brazos Co. 2; Deaf Smith Co.: Hereford (5 mi.S. W.) 3¢. FLOWER Recorps. ComposiraAE: Coneflower ?. LEGUMINOSAE: Amorpha canescens ¢, Astragalus gracilis,* Dalea sp. 6, Medi- cago sativa 49, Melilotus alba 5926, M. officinalis 2 4, Petalo- stemum sp. 2°, P. purpureum,* Psoralea sp.3¢, P. tenuiflora 2 ¢. Prepators. Cockerell (1908) found the paratype of this species struggling with a worker of Pogonomyrmex occidentalis (Hymen- optera: Formicidae ). SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Curvinomia) angustitibialis, new species (Figs 4, 15 and 21; Map 3; Table 5) This species can be differentiated from other North American Curvinomia by the coarse punctation, wide, yellow abdominal bands, and in the male by the almost unmodified hind tibia. Mate. Structure: (1) Clypeal margin produced into a median tooth; submarginal area punctate except for tooth; clypeus with longitudinal, impunctate line. (2) Metepisternum with many fine wrinkles. (3) Propodeum with coarse punctures on posterior face. (4) First m-cu several vein widths basal from first r-m. (5) Pos- terior femur three times as long as broad. (6) Posterior tibia not greatly expanded toward apex (fig. 4). (7) Tergum 1 with coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; band two rows of punctures wide, ab- sent toward center. (8) Tergum 2 with punctures like those of center of tergum 1; groove between postgradular and pregradular areas deep, its anterior face vertical, posterior face sloping gently upward. (9) Terga 3 and 4 much more finely punctate than center of tergum 1, except raised area of tergum 3 which has punctures like those of tergum 1. (10) Sternum 5 with lateral, longer part of median, apical process about as long as wide and rounded at BANDED SUBGENERA OF NOMIA IN AMERICA 293 apex, processes widely separated; Jateroapical hairy pads with hairs considerably short of meeting along midline, hairs in center of pad not darkened, pads half round in shape; median basal emargination almost three times as wide as deep, V-shaped (fig. 15). (11) Gono- coxite with apex of dorsal recurved part not bent outward and emarginate. (12) Gonostylus with plate at its base with small bumps on inner side. (13) Exposed part of penis valve little more than half as wide as long, tip narrow and pointed, inner subapical margin rounded, extreme lateral part with rounded lobe, antero- ventral process J-shaped (fig. 21). Pubescence: (14) Terga 4 to 7 with some black hairs, sometimes absent on tergum 4; tergum 3 with hairs all white. (15) Gonocoxite with dorsal recurved part with medium sized, slender hairs at tip arising from inner margin. Color; (16) Mandible translucent orange to black with red tip and red-brown base. (17) Antenna orange to red-brown below, black above. (18) Legs black to red-brown. (19) Stigma light in color like most wing veins. (20) Colored band of tergum 2 about 0.50 mm. in width; color yellow, white or green with pink, green, gold or yellow iridescence. FEMALE. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is broadened distally. (22) Clypeal margin emarginate in center. (23) Head and thorax with coarse punctures; mesoscutum with spaces between punctures minutely roughened. (24) Metepisternum with many fine wrinkles. (25) Tergum 1 with close, coarse punctures except sometimes for band of smaller punctures at base of apical, depressed, hyaline portion; band absent medially to complete and one-third width of depressed area; center of tergum not highly polished. (26) Tergum 2 with punctures same size as those of center of tergum 1 or smaller; tergum 2 not highly polished; postgradular area not depressed below level of pregradular area. (27) Terga 3 and 4 with small punctures, much smaller than those of tergum 2. Pubescence: (28) Body hairs white with some black hairs on terga 4 to 6, may be absent on tergum 4. Color: (29) Antenna red-brown below, black above. (30) Stigma darkened or light in color like most wing veins. (31) Colored band on tergum 2 about 0.50 mm. in width; color yellow or green with purple or red iridescence. VaRIATION. Little variation has been noted in the characters studied in the few specimens of this species examined. A male and a female from Taxco (23 mi. N.), Mexico, are slightly smaller in size, have narrower abdominal bands and slightly less punctate 294 THe UNIVERSITY SCIENCE BULLETIN thorax and abdomen. A male from Cuernavaca (46 mi. S.), Mex- ico, is intermediate between the Taxco male and the typical angusti- tibialis in these characters. On the basis of other characters these Mexican specimens seem to belong to this species, and the other Mexican specimens agree with those from Arizona. Type MatTertaL. Hotoryre: ¢, Southwestern Research Station (5 mi. W. Portal), Cochise Co., Arizona, August 21, 1956 (C. & M. Cazier), 5,400 ft. elev., on Melilotus alba. ALLoTYPE: @, same label data as holotype except collected on August 3 by E. Ordway. The holotype and allotype are in the collection of the American Museum of Natural History. PAratrypes: 174 and 342 from Arizona and 1é from Mexico. Arizona. Cochise Co.: Huachua Mountains, no date, U. S. Nat. Mus., 2°; Huachua Mountains, September 27, 1936 (R. H. Crandall), Univ. Ariz., ? ; Carr Canyon, Huachua Mountains, August 6, 1924 (F. P. Van Duzee), Calif. Acad. Sci., ¢; Ramsey Canyon, Huachua Mountains, September 5, 1955 _(F-G@. Werner & G. DD: Butler), Univ: Ariz.” 2 6. 2;amieave Portal, September 14, 1955 (G. E. Bohart), Utah State Univ., 9; Portal, August 12, 1944 (W. W. Jones), Univ. Calif., Riverside, 264;5 mi. W. Portal, August 21, 1956 (C. & M. Cazier), 5,400 ft. elev., on Melilotus alba, Univ. Kansas, é; 5 mi. W. Portal, August 12, 1956 (E. Ordway), 5,400 ft. elev., on Melilotus alba, Amer. Mus. Nat. Hist.,- 4; 5 mi. W. Portal, August 1, 1956"(G@. a Mer Cazier ), 5,400 ft. elev., on Melilotus alba, Amer. Mus. Nat. Hist., 34; 5 mi. W. Portal, August 2, 1956 (E. Ordway), 5,400 ft. elev., on Melilotus alba, Amer. Mus. Nat. Hist., ¢; Graham Co.: Graham Mountains, August 18, 1954 (R. M. Bohart), 3,500-4,500 ft. elev., Univ. Calif., Davis, ° ; Graham Mountains, August 15, 1953 (G. D. Butler), 5,000-6,000 ft. elev., Univ. Ariz., 9°; Pima Co.: Molino Basin, Santa Catalina Mountains, August 15, 1954 (R. M. Bohart), 5,000 ft. elev., Univ. Calif., Davis, 32; Molino Basin, Santa Catalina Mountains, September 8, 1951 (C. D. MacNeill), Univ. Calif., Berkeley, 2; Molino Basin, Santa Catalina Mountains, August 29, 1951 (C. D. MacNeill), Univ. Calif., Berkeley, 9°; Molino Basin, Santa Catalina Mountains, September 7, 1951 (C. D. MacNeill), Univ. Calif., Berkeley, 2 2° ; Molino Basin, Santa Catalina Mountains, September 9, 1951 (C. D. MacNeill), Univ. Calif., Berkeley, ¢; Santa Cruz Co.: 2.5 mi. W. Pena Blanca Lake, August 17, 1961 (J. C. Bequeart), Univ. Ariz., 26; 5.5 mi. W. Pena Blanca Lake, August 17, 1961 (Werner & Bequaert), Univ. Ariz., 9°; Sycamore Canyon, near Ruby, August 16-17, 1961 (Werner & Bequaert), BANDED SUBGENERA OF NOMIA IN AMERICA 295 Univ. Ariz., °; Madera Canyon, Santa Rita Mountains, August 9, 1947 (L. Martin), on yellow composite, Los Angeles Co. Mus., 59; Madera Canyon, Santa Rita Mountains, August 20, 1952 (L. M. Martin), Los Angeles Co. Mus., ¢; Madera Canyon, Santa Rita Mountains, August 2-4, 1959 (K. V. Krombein), U. S. Nat. Mus., 22; foot of Miller Creek?, Santa Rita Mountains, August 16, 1940 (E. C. Van Dyke), Calif. Acad. Sci., 89; foot of Miller Creek?, Santa Rita Mountains, August 16, 1940 (E. C. Van Dyke), Univ. Kansas, ¢. Mexico. Sonora: 6 mi. N. W. Cananea, August 15, 1959 (W. L. Nutting & F. G. Werner), 6,000 ft. elev., Univ. ITA 55 Bs The Mexican specimens not included in the type series are as follows: Durango: Canutillo (8 mi. S.) ?, Coyotes (4 mi. S. W. El Salto) 2¢, El Salto (6 mi. N.E.) 29; Guerrero: Taxco (23 mi. N.) @ 6; Morelos: Cuernavaca (46 mi. S.) ¢. RaNncE. This species ranges from southeastern Arizona to the state of Guerrero in Mexico (map 3). It has not been found in tropical Mexico. FLOWER Recorps. See under Type Material. SEASONAL Activity. See under Type Material. Table 5 gives a summary of the seasonal activity of the adults. Nomia (Curvinomia) foxii Dalla Torre (Figs. 5, 16 and 22; Map 2; Tables 1 and 5) Nomia punctata Fox, 1893, p. 184 (not Smith, 1875) (tax., @ ); Cockerell, 1896, p. 34 (fis. ). Nomia foxii Dalla Torre, 1896, p. 167 (new name for N. punc- tata Fox) (in part) (cat.); Cockerell, 1897, p. 20 (rec.); Cockerell, 1898b, p. 50 (in part) (tax.); Cockerell, 1898c, p. 50 (in part) (tax.); Cockerell, 1898d, p. 31 (in part) (tax., ¢ ); Cockerell, 1902, p. 812 (fs.); Cockerell, 1906a, p. 298 (in part) (rec.); Cockerell, 1906b, p. 279 (tax.); Viereck, 1906, p. 237 (rec.); Cock- erell, 1907, p. 242 (tax.); Snow, 1907, p. 186 (in part) (rec.); Cockerell, 1908, p. 333 (only ¢) (tax. 6); Cockerell, 1910a, p. 277 (only ¢ ) (tax.); Michener, 1936, p. 4 (tax.). Nomia (Paranomia) foxii: Cockerell, 1905, p. 322 (tax.); Cock- erell, 1910b, p. 290 (tax. ). Nomia mesillensis Cockerell, 1908, p. 334 (tax., ? ); Cockerell, 1910a, p. 277 (tax.); Cockerell, 1912, p. 493 (tax.); Cockerell, 1930, p. 13 (tax.); Cockerell, 1934b, p. 7 (tax.), new synonymy. Nomia (Paranomia) mesillensis: Cockerell, 1910b, p. 297 (tax.). 296 THE UNIVERSITY SCIENCE BULLETIN Nomia (Acunomia) foxii: Michener, 1951, p. 1127 (in part) (cat. ). Nomia (Acunomia) mesillensis: Michener, 1951, p. 1128 (cat.). Nomia (Paranomina) mesillensis: Cross, 1958a, p. 1266 (tax.). Nomia (Paranomina) foxii: Cross, 1958a, p. 1266 (tax.). Nomia tetrazonata: Evans & Linsley, 1960, pp. 32 and 35 (para. ); Linsley, 1960, pp. 362, 366 and 374 (para.) (misidentifica- tions ). There has been a great deal of confusion concerning the status of this species, due in part to no one’s having examined the syn- types of punctata (= foxii), and to the original description of punctata, which states the abdomen is “distinctly, sometimes coarsely punctured.” N. foxii is the least punctate North American banded Nomia. The original description also states that the abdo- men is “shining.” N. foxii is the only Curvinomia with a shining abdomen. N. punctata was described from three female specimens from the following localities: Denver, Colorado; Vega San Jose, New Mexico; and Big Stone City, South Dakota. Only the speci- men from New Mexico has been located and is what has been called mesillensis (= foxii). This form is not known to occur as far north as Denver. The two missing syntypes must be universitatis (for further details see discussion of universitatis). The specimen from New Mexico most closely agrees with the description of punctata, while universitatis most closely agrees with Fox’s (1893) var. (?). Cockerell (1906a) selected Denver as the type locality for foxii. This is an incorrect designation of the type locality, be- cause the Denver specimen probably was an wniversitatis, which is Fox’s “variety.” Cockerell later (1908) correctly selected the “New Mexico species as the true foxii” (!) (wniversitatis also occurs in New Mexico, but this was not known to him). Females that Cockerell identified as foxii prior to 1912 belong to the species now known as tetrazonata, and males he then as- signed to foxii were actually that species. However, upon finding associated males and females of foxii in 1912, he abandoned this earlier identification, assigning the males and associated females to his (1908) species mesillensis. For the females earlier called foxii, he continued to use that name, stating (Cockerell, 1912) that he believed Crawford’s (1911) moctezumae to be a synonym of foxii. N. moctezumae is now, however, known to be a synonym of tetrazonata. BANDED SUBGENERA OF NoOMIA IN AMERICA 297 The female of foxii can be differentiated from other North American Curvinomia by the shining metasoma, and the male can be differentiated by the shape of the hind tibia and lateroapical hairy pads on sternum 5. Mate. Structure: (1) Clypeal margin broadly rounded in center with wide submarginal impunctate area; clypeus with weak longi- tudinal ridge. (2) Metepisternum with many fine to a few coarse wrinkles. (3) Propodeum with medium-sized punctures on pos- terior face. (4) First m-cu several vein width basal from first r-m. (5) Posterior femur twice as long as broad. (6) Posterior tibia greatly expanded toward apex (fig. 5). (7) Tergum 1 with coarse punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band two-thirds width of depressed area to reduced to a few scattered punctures. (8) Tergum 2 with punctures smaller than in center of tergum 1; groove between postgradular and pregradular areas deep, its anterior face vertical, posterior face sloping gently upward. (9) Terga 3 and 4 much more finely punctate than in center of tergum 1. (10) Sternum 5 with lateral, longer part of median, apical process about twice as long as wide and rounded at apex, processes widely sepa- rated; lateroapical hairy pads with hairs considerably overlapping along midline, hairs in center of pad not darkened, pads crescent- shaped; median basal emargination about twice as wide as deep, V-shaped (fig. 16). (11) Gonocoxite with dorsal recurved part bent outward and irregularly rounded. (12) Gonostylus with plate at its base with small bumps on inner side. (13) Exposed part of penis valve two-thirds as wide as long, tip broad and pointed, inner subapical margin with large rounded lobe, extreme lateral part with anteriorly pointing tooth, anteroventral process consider- ably bent (fig. 22). Pubescence: (14) Terga 4 to 7 with some black hairs, sometimes absent on tergum 4; tergum 3 with hairs all white. (15) Gonocoxite with dorsal recurved part with large, stout hairs at tip arising near inner margin. Color: (16) Mandi- ble black with red tip. (17) Antenna orange to brown below, black above. (18) Legs black to red-brown with basitarsus and apex of hind tibia yellow. (19) Stigma darkened or light in color like most wing veins. (20) Colored band of tergum 2 about 0.50 mm. in width; color green with gold, red, purple, white or yellow iridescence. 298 THE UNIVERSITY SCIENCE BULLETIN FeMate. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is not broadened distally. (22) Clyp- eal margin truncate to slightly emarginate in center. (23) Head and thorax with medium-sized punctures; mesoscutum with spaces between punctures smooth. (24) Metepisternum with many fine to fewer coarse wrinkles. (25) Tergum | with scattered fine punc- tures except for band of smaller punctures at base of apical, de- pressed, hyaline portion; median width of band one-half to two- fifths width of depressed area; center of tergum highly polished. (26) Tergum 2 with punctures coarser than in center of tergum 1; highest part of tergum 2 nearly impunctate and highly polished; postgradular area not depressed below level of pregradular area. (27) Terga 3 and 4 with small punctures, much smaller than those of tergum 2. Pubescence: (28) Body hairs white with some black hairs on terga 3 to 6. Color: (29) Antenna orange to brown below, black to red-brown above. (30) Stigma darkened or light in color like most wing veins. (81) Colored band of tergum 2 about 0.50 mm. in width; color green with red-orange, purple or cream iri- descence. Type MareriaL. N. punctata (= foxii) was described from three females. One of these, from Vega San Jose, New Mexico, August 4 (C. H. Tyler Townsend), is at the Academy of Natural Sciences of Philadelphia. Vega San Jose could not be located on any map, but it is thought to refer to La Vega (a plain around San Rafael) south of the Rio San Jose in northern Valencia County. The other syntypes (Denver, Colorado, E. V. Beales and Big Stone City, South Dakota, J. F. Aldrich) could not be located, but are thought to be wniversitatis (see above and under universitatis ). Cockerell (1908) did not specifically designate the New Mexico specimen as the lectotype, but that was clearly his intent. I there- fore here designate the female from Vega San Jose, New Mexico, as the lectotype of N. foxii. The female holotype of N. mesillensis, Mesilla Valley (“towards Organ Mountains”), New Mexico, end of September (T. D. A. Cockerell), is in the U. S. National Museum, type number 55860. Rance. This species ranges from southern Colorado westward to Arizona and southward to Jalisco, Mexico (map 2). It does not occur in tropical Mexico. List oF Locatities. ARIZONA. Apache Co.: White Mountains 2°54; Cochise Co.: Apache ¢, (5 mi.S.E.) 9, (4 mi. S. W.) 9, Chiricahua National Monument 2 ¢, Dos Cabezas (16 mi. S. E.) BANDED SUBGENERA OF NOMIA IN AMERICA 299 2, Douglas 48, (8 mi. N.) 226, (18 mi. N.) 922, (20 mi. N.) Sig (2m Ne) 935 30h mie Ne 63, (45 migN:) Wolo. Bl Frida ¢, Fairbank (7 mi. W.) 2234, Huachuca Mountains (Ramsey Canyon) 3@, (6 mi. W. Montezuma Pass) 6, Pearce Ses saconmis Ne) or4 6 sRortall2i¢ = (Qemie Ne E.)) 3/0120 6, (2tanunit N.E.) @28, (5 mi. W.) 149 296, Rodeo, New Mexico (6 mi. W.) @, Sands Ranch (Dry Canyon, S. E. end Whetstone Moun- tains) ¢, Skeleton Canyon (6 mi. S.E. Apache) ¢ é, Tumer (7 mi. N.W. Fort Huachuca) ¢, Willcox 34; Coconino Co.: Flagstaff °, (15 mi N.) 43, (15-20 mi. N.) 42, (20 mi. N.) 6923, (25 mi. N.) 122 26, Mormon Lake 29, Oak Creek Canyon 1722¢, Turkey Tanks (near Cosnino) 2; Gila Co.: Cedar Creek (15 mi. W. Fort Apache) 6, Young 2; Navajo Co.: For- estdale Trading Post (9 mi. S.W. Show Low) ¢, Marsh Pass (12 mi. S. W. Kayenta) 9°, Shumway 5 ?; Santa Cruz Co.: Canelo 64, Locheil (9 mi. E.) 72, Mustang Mountains ¢, Patagonia 32984, Sonoita (10 mi. E.) 10924, Sycamore Canyon (4 mi. S. E. Ruby) 32 4; Yavapai Co.: Dewey 29 ?, Lonesome Valley (10 mi. N. E. Prescott) ¢, Mt. Mingus (6 mi. S. W.) 10 é, Prescott 89 ©2026. (6 mm NG We) By ClO iam, IN) Gy (28 ttl, ING WW) Se CatiForNtA. San Bernardino Co.: Needles ?. Cotorapo. Las Animas Co.: Hoehne (N.E.) @; Montezuma Co.: Yellow Jacket 45@ 6. New Mexico. Bernalillo Co.: Sandia Mountains 2 ¢; Catron Co.: Alma 44, (5 mi. N.) 42, Whitewater Canyon (4 mi. N. E. Glenwood) 2°; Chaves Co.: East Grand Plains (7 mi. S. E. Roswell) ¢, Roswell 212114, (5 mi. S.) ?; Colfax Co.: Koehler 22; Dona Ana Co.: Mesilla Valley (towards Organ Mountains ) 2; Eddy Co.: Hope 22; Hidalgo Co.: Lordsburg (13 mi. N. W.) OF se army) 6, (25 mina) 44 Rodeo 100 79:3 Gl mi... ) Ons Clamninite We) 2 soe mae S.) G9 34.092 ami. Ne) one die (43min N,) 9. (7 mi. S-E.)- 3; luna Cor Deming 9465) (14 mi. W.) 4; Otero Co.: Mountain Park 994; Rio Arriba Co.: El Rito 4; Roosevelt Co.: Portales 4¢; San Juan Co.: Farming- ton ¢; San Miguel Co.: Los Montoyas 19¢ é, Pecos River (Pecos ) 246, Sapello 2, San Jose ¢; Santa Fé Co.: Santa Fé 5236; Taos Co.: Embudo (6 mi. N. E.) 42; Valencia Co.: Cubero 3? ¢, Vega San Jose (near San Rafael) 2° 6. Texas. Brewster Co.: Alpine 22 ¢; Culberson Co.: ¢; Jeff Davis Co.: Fort Davis (10 mi. S.) 2324, (21 mi. W.) ¢. México. Aguascalientes: Aguascalientes (40 mi. N.) ?2¢6, Rincédn de Romos (12 km. N.) 294; Chihuahua: Ciudad Camargo (20 mi. S. W.) ¢, Catarinas (22 mi. S. E. Hidalgo 300 Tue UNIVERSITY SCIENCE BULLETIN del Parral) 36, Chihuahua (30 mi. N. W.) 64, (45 mi. N. W.) 2, Gallego (8 mi. S.) 44, Hidalgo de Parral (29 mi. E.) 16298¢, (42 mi. S.) 2, (66 mi. S.) 429194, Jiménez (18 mi. W.) 72164, (78 mi. W.) 9184, Moctezuma 2¢?, Santa Barbara 2946, (36 km. S.) ¢, Salaices 9; Coahuila: La Rosa ¢, Paila (80 mi. E. Torreén) 4, Saltillo (5 mi. N.) @; Durango: Canutillo (8 mi. S:) 9919's, Durango. 3, (14 min NE.) 63) (0 mi, Neo (23 mi. N.E.) 2276, (50 mi. S.) 2 ¢, (156 mi. N.) 20> Ememo (15 mi. N. Santa Maria del Oro) ?, Nombre de Dios 142114, (16 mi. S.) 64, San Juan del Rio 934, Torreon 934, Yerbanis 2 6; Jalisco: Encarnacion de Diaz 8 ¢ , Lagos de Moreno (15 mi. N.E.) @, San Juan de los Lagos 9428, (8 mi. S.W.) ¢; Sonora: Agua Prieta ¢; Zacatecas: Fresnillo (8 mi. S.) ¢, (9 mi. S.) 10°94, Zacatecas (34 mi.S.) 2°. TaBLE 1.—Summary of Flower Records for Nomia foxii Dalla Torre Plant Data* Records of N. foxi Number | Number | Number | Number | Number | Total FAMILY of of of col- of of number genera species | lections | females males of bees Asclepiadaceae 1 2 16 23 139 162 Capparidaceae 1 2 9 22 9 31 Compositae.... 6 6 15 16 55 71 Cruciferae..... 1 1 Dy 2 9 11 Euphorbiaceae 1 1 1 0 1 1 Leguminosae. . . 7 8 41 112 237 349 Liliaceae...... 1 1 1 1 0 1 Loasaceae..... 2 2 2 0 2 2 Malvaceae... . 1 1 1 0 1 1 Polygonaceae. . 1 1 4 9 2 11 Rubiaceae... .. 1 1 1 0 Ds 2, Solanaceae... . 1 zy 1 1 0 ] Verbenaceae. . . il 1 1 1 0 1 Zy gophyllaceae 1 1 6 4 8 12 (otal eae 26 30 101 191 465 656 * Records from the literature are included under this heading. FLowerR Recorps. ASCLEPIADACEA: Asclepias sp. 119916, A. Linaria 8964, A. subverticillata 49434. CAPPARACEAE: Cle- ome sp. 16236, C. Jonesii 2964, C. serrulata 42. ComposiraE: Baccharis sp. 10238 4, B. glutinosa 7 é , Cirsium sp. 2, Guardiola Tulocarpus 49°76, Gutierrezia sarothrae ¢°, Haplopappus sp. 2, Helianthus annuus 2 @ , Verbesina sp. 6. CructFERAE: Lepidium sp. 2°94. EupHorpiaceaE: Croton sp. é. LrcuminosaE: Crotalaria BANDED SUBGENERA OF NOMIA IN AMERICA 301 pumila @, Eysenhardtia polystachya 1027 4, Lotus Wrightii 31 9 196 4, Medicago sativa 28? , Melilotus sp. 49, M. alba 312256, M. officinalis 244, Petalostemum sp. 2 6, P. candidum 5°, Psoralea teniflora 6. LitaceaE: Nolina sp. 9. LOASACEAE: Ceral- lia sinuata ¢, Mentzelia pumila 6. MatvyaceaE: Gossypium ¢@. PotyconaceaE: Eriogonum sp. 8¢@ 6, E. Abertianum @? 6. Rusta- CEAE: Crusea subulata 24. SoLANaceaE: Solanum elaeagnifo- lium,* S. rostratum 9°. VERBENACEAE: Aloysia Wrightii 9. Zyco- PHYLLACEAE: Larrea tridentata 4? 8 ¢. The above flower records are summarized in table 1. PARASITES AND Prepators. Mr. R. C. Funk identified mites from the fourth and fifth abdominal terga of a male from Douglas (8 mi. N.), Arizona, as Glyphanoetus sp. (Acarina: Anoetidae). Mallophora bromleyi and Blepharepium secabilis (Diptera: Asili- dae) prey upon N. foxii (Linsley, 1960), listed as N. tetrazonata. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Curvinomia ) tetrazonata Cockerell This species can be distinguished from other North American Curvinomia by the narrow abdominal bands, medium- to small- sized punctures, in the female by the nonpolished metasomal terga, and in the males by the campaniform lateroapical hairy pads on sternum 5 with darkened hairs in the centers. Mate. Structure: (1) Clypeal margin broadly rounded in cen- ter with narrow submarginal impunctate area; clypeus with weak longitudinal carina. (2) Metepisternum with many fine wrinkles to wrinkles absent. (3) Propodeum with fine punctures on posterior face. (4) First m-cu several vein widths basal from first r-m. (5) Posterior femur twice as long as broad. (6) Posterior tibia greatly expanded toward apex (fig. 6). (7) Tergum | with medium-sized punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band one-third width of depressed area. (8) Tergum 2 with punctures like those of center of tergum 1; groove between postgradular and pregradular area deep, its anterior face nearly vertical, posterior face sloping gently upward or groove shallow and U-shaped. (9) Terga 3 and 4 more finely punctate than center of tergum 1, except that raised area of tergum 8 which has punctures like those of tergum 1. (10) Sternum 5 with lateral, longer part of median, apical process three times as long as wide and truncate at apex, processes widely separated; lateroapical hairy pads with hairs considerably short 302 THe UNIVERSITY SCIENCE BULLETIN of meeting along midline, hairs in center of pad much darker, pads campaniform; median, basal emargination about twice as wide as deep, V-shaped (fig. 17). (11) Gonocoxite with dorsal recurved part bent outward and rounded. (12) Gonostylus with plate at its base papillose on inner side. (13) Exposed part of penis valve two-thirds as wide as long, tip broad and pointed, inner subapical margin with elongate lobe, extreme lateral part with anteriorly pointing tooth, anteroventral process slightly bent (fig. 23). Pu- bescence: (14) Terga 5 to 7 with some black hairs; terga 3 and 4 with hairs all white. (15) Gonocoxite with dorsal recurved part with large, long, slender hairs at tip arising from well behind inner margin. Color: (16) Mandible orange to black with red tip. (17) Antenna orange to brown below, black to red-brown above. (18) Legs black to red-brown with basal part of hind basitarsus and hind tibial lobe yellow. (19) Stigma darkened or light in color like most wing veins. (20) Colored band on tergum 2 about 0.33 mm. in width; color green with purple, gold or pink iridescence. Femate. Structure: (21) Labrum with median longitudinal ridge, upper margin of which is not broadened distally. (22) Clypeal margin emarginate in center. (23) Head and thorax with fine punctures; mesoscutum with spaces between punctures smooth to minutely roughened. (24) Metepisternum with many fine wrinkles. (25) Tergum 1 with close, fine to minute scattered punctures except for band of smaller punctures at base of apical, depressed, hyaline portion; median width of band one-half to two- fifths width of depressed area; center of tergum not highly polished. (26) Tergum 2 with punctures like those of center of tergum | or smaller, apical higher portion sometimes nearly impunctate; ter- gum 2 not highly polished; postgradular area not depressed below level of pregradular area. (27) Terga 3 and 4 with fine to minute punctures like those of tergum 2 or finer. Pubescence: (28) Body hairs white with some black hairs on terga 4 to 6, may be absent on tergum 4. Color: (29) Antenna orange to red-brown below, black to red-brown above. (30) Stigma darkened or light in color like most wing veins. (31) Colored band on tergum 2 about 0.33 mm. in width; color green with red, red-orange, purple or pink iridescence. Nomia (Curvinomia ) tetrazonata tetrazonata Cockerell (Figs. 17 and 23; Map 1; Tables 2 and 5) Nomia foxii: Cockerell, 1898a, p. 311 (fls.); Cockerell 1898b, p. BANDED SUBGENERA OF NOMIA IN AMERICA 303 50 (in part) (tax.); Cockerell, 1898c, p. 50 (in part) (tax.); Cock- erell, 1906a, p. 298 (in part) (rec.); Snow, 1907, p. 136 (in part) | (rec. ); Cockerell, 1908, p. 334 (only @) (tax., 2); Cockerell, 191l0a;- ps 276) (only 2) (tax); ‘Cockerelly 191125 pp. 493° (tax. ):; Cockerell, 1930, p. 13 (tax.) (misidentifications ). Nomia (Paranomia) tetrazonata Cockerell, 1910b, p. 297 (tax., 8). Nomia moctezumae Crawford, 1911, p. 618 (tax., 2 ); Cockerell, 1912, p. 493 (tax.); Cockerell, 1923, p. 103 (rec.); Cockerell, 1930, p. 13 (tax. ). Nomia tetrazonata: Cockerell, 1930, p. 13 (tax.); Michener, 1936, p. 4 (tax.); Michener, 1937, p. 319 (rec.); Linsley, 1962, p. 79 (iis. ). Nomia californiensis Michener, 1937, p. 318 (tax., 9? ), new synonymy. Nomia (Curvinomia) californiensis:; Michener, 1944, p. 251 (tax.); Michener, 1951, p. 1127 (cat.). Nomia (Acunomia) foxii: Michener, 1951, p. 1127 (in part) (cat.) (misidentification ). Nomia (Acunomia) tetrazonata: Michener, 1951, p. 1128 (cat.) Nomia (Paranomina) californiensis: Cross, 1958a, p. 1266 (tax. ) Nomia (Paranomina) tetrazonata: Cross, 1958a, p. 1266 (tax. ) Cockerell was misled by Fox's description of punctata (= foxii), and considered the female t. tetrazonata as foxii (see discussion under foxii for further details). N. californiensis is a t. tetrazonata that has lost the bright color of the abdominal bands; such loss occurs when a specimen becomes wet. N. t. tetrazonata can be separated from t. uwvaldensis by the char- acters given below. . . . Mate. (8). Tergum 2 with groove anterior to center, punctate part of postgradular area about three times as wide as colored band; groove deep, anterior face vertical, posterior face sloping gently upward; depressed area strongly and abruptly depressed, not much wider than colored band; terga 2 to 4 with punctures deep and sharp. FEMALE. (25) Tergum | with close, fine punctures not much finer than those of mesoscutum. (26) Tergum 2 with punctures like those of tergum 1, apical, higher portion with many punctures. (27) Terga 3 and 4 with fine, close punctures like those of tergum 1 or finer. 304 THE UNIVERSITY SCIENCE BULLETIN VARIATION. Little variation has been noted in the characters studied even though a large number of specimens of this subspecies were examined. The few characters that separate t. tetrazonata from t. uvaldensis are constant, and no intergrades have been found, but few specimens have been examined from the area where the ranges of these subspecies meet. Type MartertaL. The male holotype of N. tetrazonata, Globe, Arizona, July 25, 1892 (W. H. Ashmead collection), is in the U. S. National Museum, type number 12988. There were two male paratypes with these label data; one of these is in the U. S. National Museum. Another male paratype in the same museum is from the Bradshaw Mountains, Arizona, June 21, 1892 (W. H. Ashmead collection). The female holotype of californiensis, Cushinbury Springs (north side San Bernardino Mountains ), California, August 19, 1933 (C. D. Michener), is at the University of Kansas, type number 5933. The female holotype of moctezuwmae, Guaymas, Sonora, Mexico (L. O. Howard), is in the U. S. National Museum, type number 13456. Rance. This subspecies ranges from Nevada and southeastern California to central New Mexico and southward to Sonora and Baja California (map 1). List oF Loca.itres. ARIZONA. Apache Co.: Chinle ¢; Co- chise Co.: Benson 7 ? 20¢, (5 mi. W.) 54, Douglas (8 mi. E.) ¢, (17 mi. E.) 26, Fort Huachuca (11 mi. N. W.) 22, Kansas Settle- ment 2°, Portal 9°23, (2 mi? N:E.) 936, (20 mi Nee (5 mi. E.) 1029136, (10 mi. E.) 6, (15 mi. W.) 2¢, San Bernar- dino Ranch (15 mi. E. Douglas) ¢, Sands Ranch (Dry Canyon, S.E. end Whetstone Mountains) ?, Texas Pass ¢, Tombstone 928, (12 mi. W.) 324, Willcox (4.2 mi. W.) 2 &; Coconino Co.: Oak Creek Canyon ? ¢, Sedona ? 6, Walnut (= Winona) 24; Gila Co.: Claypool (5 mi. W. Globe) ¢, Gila River Canyon (near Christmas) 48, Globe 2958, (12 mi. N.) ¢?, Roosevelt 172; Graham Co.: Aravaipa Creek ?, Fort Thomas 3¢, Geronimo (9 mi. N. W.) @, Graham Mountains (15 mi. N. Wilcox) @, Roosevelt Lake (E. end) 13¢, Safford ¢; Maricopa Co.: Cavecreek 92 ¢, Chandler ¢, Gila Bend (25 mi. E) ¢, Glendale 9, Granite Reef Dam (10 mi. N.E. Mesa) ¢, Mesa 9, (25 mi. E.) 4, Phoenix 52, Tempe @, Wickenberg ¢; Pina Co.: Arivaca ¢, Baboquivari Canyon 5¢, Brown’s Canyon (Baboquivari Mountains) 34, Con- tinental ¢, Elkhorn Ranch (N. E. end Baboquivari Mountains) ¢, Mount Lemmon Road (Santa Catalina Mountains) 32, Organ BANDED SUBGENERA OF NOMIA IN AMERICA 305 Pipe Cactus National Monument 29, Ouijotoa (30 mi. E) @, Sabino Canyon (Santa Catalina Mountains) 1392¢, Sahuarita — é, San Vicente 24, Sells 2, Tucson 2891076, (10 mi. E.) ¢, (18 mi. S.) g, Tucson Mountains ¢6¢; Pinal Co.: Boyce Thomp- son Southwestern Arboretum 38 ¢ , Coolidge (17 mi. N. W.) 9°24, Floy (11 mi. S.W.) ?8¢, Florence Junction ¢, Oracle 492 ¢, Pepper Sauce Canyon (Santa Catalina Mountains) ¢, Picacho Pass 209, Santa Catalina Mountains ¢, Toltec (10 mi. S.) ?; Santa Cruz Co.: Canelo 2? ¢, Madera Canyon (Santa Rita Mountains ) 6, Patagonia 29, Santa Rita Mountains 139144; Yavapai Co.: Ash Fork?, Bradshaw Mountainsé, Campe Verde 32, Congress Junction ¢ ¢, Cottonwood (8 mi. N. E.) 4?, Dewey 924, Jerome (12 mi. S.) 9, Mt. Mingus (6 mi. S.W.) 38, Seligman ¢, (7 mi. W.) 2°, Yarnell to Prescott ¢; Yuma Co.: Palm Canyon (Kofa Miotmtams)2 a, avumia 2.6. (lovmi Ni) 205 (30 min IN] o. Cauirornia. Fresno Co.: Coalinga (22 mi. W.) 72; Imperial Co.: Beal’s Well (18 mi. N. E. Calipatria) 9, (46 mi. N.E.) 9, Calexico 6, El Centro (E. of) @; Inyo Co.: Antelope Springs (8 mi. S. W. Deep Springs) 5924, Big Pine Creek (5 mi. S. W. Big Pine) ¢@, Deep Springs (22 mi. E. Bishop) 2¢?, Inyo Moun- tains (12 mi. E. Lone Pine) @, Mazourka Canyon (Inyo Moun- tains) 24, Surprise Canyon (Panamint Mountains) 54, West- gard Pass (4 mi. W.) 62 4, (7 mi. W.) 6¢@; Riverside Co:: Andreas Canyon (4 mi. S. Palm Springs) ¢@, Cathedral City ¢, Coachella Valley (5 mi. E. Palm Springs) ¢, Gavilan (4 mi. S. E. of E. end Lake Mathews) ¢?, Hayfield Road (21 mi. N.E. of N. E. end Salton Sea) 42, Palm Desert ? ¢, Palm Springs 9°46, Palm Springs Station (7 mi. N.W. Palm Springs) ¢, Pinyon Flat (16 mi. S.W. Palm Springs) ¢, Riverside 2° 6, Soboba Hot Springs ¢°, Taquitz Canyon (2 mi. S.E. Palm Springs) °, Whitewater Canyon (8 mi. N. W. Palm Springs) ¢; San Bernar- dino Co.: Cushinbury Springs (N. side San Bernardino Moun- tains) °, Earp (10 mi. N.) ?, San Bernardino ¢, Victorville ¢ ; San Diego Co.: Anza-Borego Desert State Park ?, Borego 956, Box Canyon (15 mi. S. W. Borego) ¢, Jacumba 2?2¢4, Mason Valley (16 mi. S.W. Borego) ¢, Palm Canyon (8 mi. N. W. Borego) ?. Nervapa. Curchill Co.: Hazen (3 mi. W.) ¢; Douglas Co.: Topaz Lake ¢ ¢; Lyon Co.: Dayton 9, Yerington 35 9 25 ¢; Washoe Co.: Sparks ¢, Sutcliffe 29. New Mexico. Dona Ana Co.: Las Cruces ?, Rincon ¢; Hidalgo Co.: Lordsburg (10 mi. N. E.) 2, Rodeo 7¢, (14 mi. N.) 2, (18 mi. N.) 29; Santa Fé 306 THE UNIVERSITY SCIENCE BULLETIN Co.: Santa Fe @. Wexas: Ell Paso €o.:* El Pasov2 Gee Uae Washington Co.: Anderson’s Ranch (8 mi. N. Hurricane) ¢, St. George 446, Veyo 2? 6. Mexico. Baja California: El Mayor (2 mi. S.) ¢, San Agustin (20 mi. W.) ?; Baja California Terri- torio Sur: Canipolé ¢ 2, Coyote Cove (Bahia Concepcion) 2 4, La Paz ¢, Sierra de la Laguna (Big Canyon?) ¢; Sonora: Alamos (10 mi. W.) 34, Guaymas 42, (16 mi. N.) 26, (37 mi. N.) 92 6. TaBLE 2.—Summary of Flower Records for Nomia t. tetrazonata Cockerell Plant Data Records of N. t. tetrazonata Number | Number | Number | Number | Number Total FAMILY of of of col- of of number genera species | lections | females males of bees Asclepiadaceae 2 2 4 4 2 6 Boraginaceae. . 1 1 1 0 1 1 Capparidaceae 3 3 4 6 0 6 Compositae.... 7 7 11 5 13 18 Convolvulaceae 1 1 2 6 2 8 Cruciferae... .. 1 1 1 0 1 1 Euphorbiaceae 2 2 6 | 2 7 9 Gramineae... . 1 1 ieee 1 0 1 Labiatae...... 1 1 1 0 1 1 Leguminosae.. . 8 8 30 4] 51 92 Liliaceae... .. 1 1 1 @) 1 1 Malvaceae... . 1 1 2 2 0 2) Myrtaceae... 1 1 1 0 2 2 Nyctaginaceae | 2 2 3 2 1 3 Polygonaceae. . Dy 5 8) 23 10 33 Rhamnaceae... 1 1 ] 23 0) 23 Sapindaceae. . . 1 1 1 1 1 2 Scrophularia- CEG sb 506 1 il 1 1 0) 1 Solanaceae... . 1 2 8 16 0 16 Vitaceae..... | 1 1 1 0 3 3 Zy gophyl- | laceae..... D 2 Saeee| 15 9 24 Mortals 41 45 107 148 105 253 FLower Recorps. ASCLEPIADACEAE: Asclepias sp. 42 6, Phili- bertia heterophylla ¢. BORAGINACEAE: Heliotropium sp. 6. Cap- PARIDACEAE: Cleome sp. 22, Cleomella longipes 9°, Wislizenia refracta 39. ComposiraE: Baccharis sp. 236, B. sarothroides 2, Chrysothamnus sp.?, Encelia sp.2, E. farinosa 2¢, Gutier- rezia sp. 6, G. Sarothrae 2, Heterotheca sp. 4 6, Hymenothrix Wis- lizeni ¢, Verbesina sp. 26. CONVOLVULACEAE: Cuscuta sp. 2 ¢, C. umbellata 69. CRUCIFERAE: Lepidium sp. 6. EUPHORBIACEAE: BANDED SUBGENERA OF NOMIA IN AMERICA 307 Croton californicus 944, Euphorbia sp. 934. GRAMINEAE: Zea Mays ¢. Lasratar: Hyptis Emoryi ¢. LEGUMINOSAE: Acacia sp.49148, Cercidium sp. ¢ 6, C. floridum 946, Dalea sp.52 6, Glycine Max ¢, Lotus scoparius 8°, Medicago sativa 29264, Melilotus sp.° 6, M. alba 142104, Prosopis sp. 42 é, Prosopis sp./Acacia sp. 17 4. Litracear: Nolina parryi 6. MALVACEAE: Gos- sypium sp. 29. MyrraceaE: Melaleuca sp. 246. NyCTAGINACEAE: Allionia incarnata ¢ 6, Boerhaavia sp. 2. POLYGONACEAE: Anti- gonon sp. @ 6, Eriogonum sp. 11984, E. fasciculatum 9°, E. heermannti 59 6, E. inflatum 49°, E. nodosum ?. RHAMNACEAE: Condalia sp. 234. SaprnpacEAE: Sapindus sp. ¢, S. Sponaria 2. SCROPHULARIACEAE: Penstemon Palmeri 2. SOLANACEAE: SO- lanum sp. 2, S. elaeagnifolium 149, S. rostratum 9°. V1ITACEAE: Cissus trifoliata 38. ZycopHyLLACEAE: Larrea tridentata 12°92, Kallstroemia grandiflora 3 ° . The above flower records are summarized in table 2. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Curvinomia ) tetrazonata uvaldensis Cockerell, new combination (Fig. 6; Map 1; Table 5) Nomia (Paranomia) uvaldensis Cockerell, 1930, p. 13 (tax., @ ). Nomia (Paranomia) zabriskii Cockerell & Blair, 1930, p. 18 (tax., ¢ ), new synonymy. Nomia zabriskii: Blair, 1930, p. 17 (tax.). Nomia parksi Cockerell, 1934b, p. 7 (tax., 2? ), new synonymy. Nomia uvaldensis: Cockerell, 1934b, p. 7 (tax. ). Nomia (Acunomia) uvaldensis: Michener, 1951, p. 1128 (cat.). Nomia (Acunomia) zabriskii: Michener, 1951, p. 1128 (cat.). Nomia (Acunomia) parksi: Michener, 1951, p. 1128 (cat.). Nomia (Acunomia) foxii: Wille, 1958, p. 544 (heart) (:misidenti- fication ). Nomia (Paranomina) parksi: Cross, 1958a, p. 1266 (tax.). Nomia (Paranomina) uvaldensis; Cross, 1958a, p. 1266 (tax.). Nomia (Paranomina) zabriskii: Cross, 1958a, p. 1266 (tax. ). N. uvaldensis and zabriskii are both available names for this sub- species. Since uvaldensis has line preference and its holotype is in better condition than that of zabriskii, it should be used. N. parksi is the same as uvaldensis. 308 THE UNIVERSITY SCIENCE BULLETIN N. t. uvaldensis can be separated from t. tetrazonata by the char- acters given below. Mate. (8) Tergum 2 with groove in center, punctate part of postgradular area about twice as wide as colored band; groove shallow and U-shaped; depressed area slightly and gradually de- pressed, much wider than colored band; terga 2 to 4 with punctures shallow and some punctures not sharp. FEMALE. (25) Tergum 1 with scattered, minute punctures much finer than those of mesoscutum. (26) Tergum 2 with punctures like those of tergum 1, apical higher portion with few scattered punctures. (27) Terga 3 and 4 with minute, scattered punctures like those of tergum 1 or finer. VaRIATION. Little variation has been noted in the few specimens of this subspecies examined. No intergrades have been found be- tween ¢t. uvaldensis and t. tetrazonata. This form may be found to be a good species. Since the differences between the two sub- species of tetrazonata are much less than between other species of North American banded Nomia, and since the distributions are allopatric with few specimens known from near the meeting of their ranges, it seems best to consider them as the same species. Type Matertat. Cockerell gave the following data on the holo- types of both wvaldensis (female) and zabriskii (male): Uvalde, Texas, September 30, 1899 (J. L. Zabriskie collection, presented by H. F. Wickman). These specimens also bear the date June 18-20. From the label data it seems that these specimens were collected by H. F. Wickman on June 18-20, and received by J. L. Zabriskie on September 30, 1899. .The latest date I have seen on any other specimen from Texas is August 25. The holotype of uwwvaldensis has the following parts missing: right front leg, left middle leg beyond the middle of the femur, right middle leg beyond femur, antennae, mouth parts and mandibles. The holotype of zabriskii has the following parts missing: antennae, two tarsal segments of right front leg, four tarsal segments of left hind leg, hidden sterna and genital capsule. This specimen has also lost the color of the abdominal bands and has the integument discolored. Cockerell mentions the antennae as being missing on these specimens, but the rest of the damage must have occurred since their description. In zabriskii this might have happened when Blair (1930) dissected the genitalia and hidden sterna. These holotypes along with that of parksi, female, Bexar County, Texas, May 27, 1931 (H. B. Parks), are in the American Museum of Natural History. BANDED SUBGENERA OF NOMIA IN AMERICA 309 Rance. This subspecies ranges from eastern New Mexico and Texas southward to the States of Durango and San Louis Potosi in Mexico (map 1). List oF Locatitres. New Mexico. Guadalupe Co.: Santa Rosa 26; Quay Co.: Tucumcari 9. Texas. Bexar Co. 2¢: Fort Sam Houston ¢; Brewster Co.: Big Bend National Park ¢, Santa Elena Canyon (Big Bend National Park) 102; Dawson Co.: Lamesa ¢; Dimmit Co.: Carrizo Springs 3914¢; Hidalgo Co. 234: Hidalgo 2966; Jim Wells Co.: Alfred 32396; Maverick Co.: Quemado @ ¢; Pecos Co.: Fort Stockton ¢; San Patricio Co.: Lake Corpus Christi ¢; Starr Co. 6?2¢,; Rio Grande City (5 mi. E.) 9; Uvalde Co.: Uvalde 926; Webb Co.: Laredo 6, (21 mi.N.) 2, (18 mi.N.) @ 6; Zavala Co.: Nueces River ¢?. Mexico. Chihuahua: Chihuahua ?; Coahuila: Gloria (S. of Mon- clova) @, Piedras Negras (192 km. S.) 2; Durango: La Loma (15 mi. S.W. Torreon) 4%, Pedriceha 2 6; Nuevo Leon: Monte- morelos 46, Vallecillo 425¢; San Luis Potosi: Santa Domingo ¢. FLOWER Recorps. CompositaE: Baccharis glutinosa 9. Lasia- TAE: Teucrium cubense 5? 6. LecumMINosAE: Acacia sp. é, A. Greggii 32°39 6, Cercidium texanum 2, Melilotus alba 2 , Park- insonia aculeata 22 ¢, Prosopis sp. 2°, P. juliflora ¢. SOLANACEAE: Solanum elaeagnifolium °. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Subgenus Acunomia Cockerell Nomia (Acunomia) Cockerell, 1930, p. 11 (type species: N. nortoni Cresson); Sandhouse, 1943, p. 522; Michener, 1944, p. 251; Michener, 1951, p. 1127; Cross, 1958a, p. 1263. When first proposed by Cockerell, Acunomia included nortoni and the melanderi complex. Michener (1951) included all the metallic-banded Nomia of the United States in Acunomia. Cross (1958) placed robinsoni in Acunomia, and considered Curvinomia (= Paranomina) as a distinct subgenus with the same species as are included here. The characters that separate Acunomia from the North American Curvinomia are given below. Mate. Antenna tapering and reaching beyond thorax, last fla- gellar segment pointed or rounded. Posterior tibia produced an- teriorly and distally into a large lobe hiding all or almost all of 13—1367 310 THE UNIVERSITY SCIENCE BULLETIN tibial spurs in side view; tibia not triangular in shape as seen from side (fig. 7). Tergum 2 (sometimes 1) to 5 with colored bands. Sternum 3 with ventrolateral process. Sternum 5 with setae in row behind lateroapical hairy pads completely fused except in robinsoni; median apical process arising from median longitudinal ridge (fig. 24). Penis valve with ventral opening; gonostylus with small, inward pointing, hooklike process at base (fig. 30). FremMate. Middle tibial spur with few, uneven, coarse teeth along apical third. Tergum 2 (sometimes 1) to 5 with metallic bands. Nomia ( Acunomia ) nortoni Cresson This is the largest species of Acunomia, and can be easily dis- tinguished from any other North American bee by its large size and the integumental colored abdominal bands. Mae. Structure: (1) Mandible with large rounded lobe on posterior margin. (2) Clypeus with bare part impunctate, nearly flat or slightly depressed in center or with weak longitudinal carina; subapical area slightly depressed. (3) Last flagellar segment acutely pointed, one and one-half to two times as long as second flagellar segment. (4) Pronotum without carina on_ posterior, dorsal edge. (5) Scutellum deeply to slightly depressed in center, lateral higher parts less densely punctate than surrounding area. (6) Metepisternum coarsely wrinkled. (7) Propoedum with an- terior groove widened to narrowed medially. (8) Tegula emargi- nate or straight on outer edge, subtruncate posteriorly. (9) Front wing length (see section on methods) about 11.2 mm. (range: 10.6-12.2, one specimen 10.0). (10) Middle tibial spur straight. (11)Posterior femur as broad as long, large basal and lateral areas nearly impunctate and shining. (12) Posterior tibia produced anteriorly and distally into large rounded lobe which hides tibial spurs in side view (fig. 7). (13) Metasomal tergum 1 coarsely punctate except for band of minute punctures at base of apical, depressed, hyaline portion; median width of band varies from six rows of punctures to few scattered punctures, width of band re- duced laterally. (14) Terga 2 to 4 with few coarse punctures like those on tergum | on raised part of postgradular area; otherwise with small, close punctures except on colored bands. (15) Sternum 3 with ventrolateral, toothlike process twice as wide as high. (16) Ventroapical, median part of gonobase not produced. (17) Penis valve with large triangular ventral opening; valve produced lat- erally into two projections, the outer acute, the inner narrowly BANDED SUBGENERA OF NOMIA IN AMERICA Olel: rounded; tip robust and pointed (fig. 30). Pubescence: (18) Face, including upper third of clypeus and lateral part of scape, covered with short, appressed, plumose, yellow or yellow-white hairs. (19) Longest hairs of face one-half to two-thirds as long as scape. (20) Anterior margin of genal area, along eye, bare or with short, appressed, plumose hairs. (21) Hair of anterior part of mesoscutum, of posterior lobe of pronotum, and anterior part of mesepisternum yellow-white to white. (22) Median meso- scutal line and notaulus obscured by hairs except posteriorly. (23) Metanotum with anterior forth to half obscured by short, plumose hairs. (24) Tegula covered with fine simple hairs posteri- orly. (25) Propodeal hairs white. (26) Longest hairs of tergum 2 as long as to twice as long as lateral width of colored band. (27) Tergum 4 usually and sometimes tergum 3 with some black hairs, terga 2 with hairs all white. Color: (28) Mandible translu- cent yellow, orange, or red-brown, with subapical red area. (29) First seven flagellar segments yellow to orange below, other seg- ments orange below. (30) Legs black to red-brown except tarsi and undersides and apices of tibiae, which are yellow to orange; last four tarsal segments of middle and hind legs brown above. (31) Stigma brown, about same color as most veins; vein R, inter- section of M and Cu, and distal end of V black to dark brown. (32) Terga 2 to 5 with apical colored bands; band of tergum 5 absent laterally; band color variable, blue, green or yellow with purple, pink or gold iridescence. Femate. Structure: (33) Clypeal margin truncate to slightly emarginate medially; part overhanding base of labrum not hyaline; longitudinal carina distinct to weak; clypeus evenly punctate. (34) Face with elevated portion below antennae impunctate or with few punctures. (35) Pronotum without carina on posterior, dorsal edge. (36) Mesoscutum and scutellum with scattered punc- tures; scutellum slightly depressed or flat in center. (37) Mes- episternum with coarse wrinkles. (38) Propodeum with anterior groove sometimes widened medially. (39) Tegula slightly emar- ginate or straight on outer edge, subtruncate posteriorly. (40) Front wing length (see section on methods) about 9.9 mm. (range: 9.2-10.7, one specimen 8.4). (41) Metasomal tergum 1 with medium punctures at sides, becoming smaller and more scattered towards center except small oval center area usually with much closer punctures, and band of minute punctures at base of apical, depressed, hyaline portion; median width of band one-half to S12 THE UNIVERSITY SCIENCE BULLETIN two-thirds width of hyaline portion. (42) Tergum 2 with medium punctures at sides of basal area, becoming much finer toward center. (43) Tergum 3 with gradular ridge distinct and with few scattered punctures. Pubescence: (44) Longest hairs of face as long as or longer than distance from antenna to inner eye margin. (45) Anterior margin of genal area, along eye, bare or with short, appressed, plumose hairs. (46) Mesoscutal hairs mostly dark. (47) Notaulus obscured; median mesoscutal line visible or ob- scured anteriorly by long hairs. (48) Metanotum completely to its anterior third obscured by short, plumose hairs; longer black hairs present or absent. (49) Tegula covered with fine simple hairs posteriorly. (50) Tergum 1 with lateral part of depressed, hyaline portion obscured by subappressed mat of hairs. (51) Terga 3, 4 and sometimes 2 with some black hairs. Color: (52) Antenna light brown below. (53) Stigma brown, about same color as most veins; vein R, sometimes intersection of M and Cu, and distal end of V black to dark brown. (54) Legs dark brown. (55) Terga 2 to 5 with apical colored bands; band color variable, yel- low or green with gold, pink or yellow iridescence. Nomia (Acunomia ) nortoni nortoni Cresson (Figs. 7, 24 and 30; Diagrams a and b; Map 4; Tables 3 and 5) Nomia nortoni Cresson, 1868, p. 385 (tax., 6, 2); Cresson, 1872, p.. 256 (rec.); Gresson, 1875, p. 722: (rec. )-) Pattomyalis79: p. 366 (in part) (tax.); Snow, 1880, p. 95 (rec.); Cragin, 1886, p. 211 (rec.); Cresson, 1887, p. 294 (cat.); Robertson, 1892° p- 105 (fls.);,Cockerell, 1893, p. 339 (rec.); Fox, 1893, p. 134 (tax. ); Dalla Torre, 1896, p. 168 (in part) (cat.); Cockerell, 1898b, p. 50 (tax.); Cockerell, 1898c, p. 50 (tax.); Cockerell, 1898d, p. 32 (tax.); Birkman, 1899, p. 244 (rec.); Bridwell, 1899, p. 210 (tec.); Cockerell, 1899a, p. 4 (rec.); Cockerell, 1899b, p. 7 (cat.); Cock- erell, 1906a, p. 298 (rec.); Cockerell, 1906b, p. 279 (tax.); Swenk, 1907, p. 294 (rec.); Cockerell, 1914, p. 411 (fls.); Cresson; 1916: p. 108 (tax. ); Cockerell, 1917, p. 190 (rec.); High, 1921, p: 307 (biol.); Blair, 19380, pp. 7 and 17 (tax.); Cockerell, 1934b, p. 7 (biol. ); Sandhouse, 1943, p. 522 (tax. ). Nomia cressoni: Vachal, 1897, p. 92 (in part) (tax., 9) (mis- identification ). Paranomia nortoni: Cockerell, 1900, p. 407 (rec.); Cockerell, 1904, p. 42 (tax.); Robertson, 1925, p. 434 (fls.); Robertson, 1928, pp. 10, 65, 120, 123, 130, 134, 148 and 198 (fls.); Brimley, 1938, p: 407 (mee). BANDED SUBGENERA OF NOMIA IN AMERICA 313 Nomia nortoni var. plebeia Cockerell, 1910a, p. 276 (tax.). Nomia (Paranomia) nortoni: Cockerell, 1910b, p. 290 (in part) (lax): Nomis [sic] nortonii: Parks, 1928, p. 263 (biol. ). Nomia (Acunomia) nortoni: Cockerell, 1930, p. 11 (tax.); Mich- ener, 1951, p. 1127 (in part.) (cat.); Wille, 1956, p. 450 (muscu- lature ); Cross, 1958a, p. 1266 (tax.); Krombein, 1958, p. 233 (cat.); Mitchell, 1960, p. 506 (tax., ¢, 92 ). Nomia (Acunomia ) nortoni plebeia: Cockerell, 1930, p. 16 (tax.). Nomia (Acunomia) nortoni var. plebia: Michener, 1951, p. 1128 (lapsus calami ) (cat.). In his description of this species Cresson said the male had three hind tibial spines. The male syntype has the hind tibial spurs on one leg obscured by glue, and on the other leg some of the hair has become matted together giving the impression of a third tibial spur on that leg. Cresson’s error led Westwood to describe cressoni. Mate. (8) Tegula emarginate, occasionally slightly so, on outer edge. (12) Hind tibial lobe with raised area of outer surface limited to basal half, raised area not forming distinct ridge; lobe at least 1.8 mm. wide (fig. 7). (19) Hairs short, longest hairs of face about half as long as scape. (20) Anterior margin of genal area, along eye, bare. (26) Longest hairs of tergum 2 about as long as lateral width of colored band. FemMate. (37) Scutellum slightly depressed or flat in center. (40) Tegula slightly emarginate on outer edge. (45) Hairs short, longest hairs of face about as long as distance from antenna to inner eye margin. (46) Anterior margin of genal area, along eye, bare. (48) Median mesoscutal line not obscured anteriorly by long hairs. VaRIATION. In spite of its wide range, this subspecies exhibits comparatively little interpopulational variation in the characters studied. The variation that does occur is considered in the de- scriptions above and is not distributed in a clinal fashion, but occurs randomly throughout the range. Type Matertat. N. nortoni was described from three females and a male: Female, New Mexico, summer of 1867 (Samuel Lewis), Academy of Natural Sciences of Philadelphia, type number 2135; female, Colorado (James Ridings), Academy of Natural Sci- ences of Philadelphia, type number 2535.3; male, same data with, type number 2535.2; female, Kansas (Edward Norton collection), 314 THE UNIVERSITY SCIENCE BULLETIN location now unknown. In 1916 Cresson designated the New Mex- ico specimen as the lectotype. The female holotype of N. nortoni var. plebeia, Fedor, Texas, June 18, 1905 (G. Birkman), is in the U. S. National Museum, type number 55861. Rance. This subspecies ranges from Pennsylvania to the Rocky Mountains and south into southern Texas and Florida (map 4). West of the Rocky Mountains it is replaced by a closely related, smaller species, N. melanderi. An apparent exception to this state- ment is one specimen of nortoni collected in southern Idaho; it is presumably mislabeled. It is unlikely that this species occurs in southern Idaho since it has never been found there by G. E. Bohart or G. F. Knowlton who have collected extensively in that area. List oF Locauitimes. ALaBaMa. Mobile Co.: Mobile 9@ ¢; Dallas Co.: Selma ? ¢. Arkansas. Ashley Co.: Wilmot (6 mi. E.) 4; Sevier Co.: De Queen ¢. CoLtorapo. Clear Creek Co.: Mt. Evans (Bear Creek Canyon) ¢; Custer Co.: Westcliffe; * El] Paso Co.: Manitou ¢. Fiorma. Alachua Co.: Gainesville 32524; Calhoun Co.: Blountstown ¢?; Jackson Co.: Cottondale 2; Jefferson Co.: Monticello 72, Lamont ¢; Levy Co.: Morris- ton 1199&é4,; Madison Co. ¢; Marion Co.: Dunnellon @; Pasco Co.: Zephyrhills 209 ¢; Sumter Co. ¢; Suwannee Co.: Bradford 3954, Suwannee Springs ¢; Taylor Co.: Perry 6; County ?: Swain 48. Gerorcia. Clarke Co.: Athens 9; Crawford Co.: Roberta ¢@; Decatur Co.: Bainbridge @ ¢, Butlers Ferry 4, Spring Creek 2?; Dougherty Co.: Albany 9¢; Jefferson Co.: Wadley @; Thomas Co.: Thomasville ¢. IpaHo. Caribou Co.: Soda Springs @. Ittrnors. Adams Co. ¢; Alexander Co.: Cache @; Champaign Co.: Urbana ?; Macoupin Co.: Carlinville; * Woolford Co.: Kappa @. Kansas. Allen Co. ¢; Barber Co.; * Cheyenne Co. 24; Douglas Co. 46?99¢: Baldwin City 29 6, Lawrence ¢; Finney Co.: Garden City ?7¢; Kearny Co. 3924; Neosho Co. é; Pottawatomie Co.: Onaga ¢; Riley Co. 14?12¢: Manhattan @; Sumner Co. ? ¢; Wallace Co. 109: Wallace 29 104; Wilson Co. ¢. lLourstana. Ascension Co.: Darrow 2 2; Cameron Co.: Creole 1729188; East Baton Rouge Co.: Baton Rouge 2234; Orleans Co.: New Orleans ¢; Point Coupee Co.: New Roads ¢1?. Maryianp. Allegany Co.: Cumberland @. Mississippi. Bolivar Co.: Benoit ¢; Noxubee Co.: Shuqualak 11¢. Mussourt. Miller Co.: Lake Ozark ¢; Taney Co.: Bran- son ¢. Nepraska. Cass Co.: Union ¢?; Clay Co.: Harvard 2; Douglas Co.: Omaha ?; Dundy Co.: Haigler ¢; Lancaster BANDED SUBGENERA OF NOMIA IN AMERICA SB) Co.: Lincoln 6? 6; Otoe Co.: Nebraska City 13954; Scotts Bluff Co.: Mitchell 42?10¢. Netw Mexico. Chaves Co.: Ros- well 39846; San Miguel Co.: Las Vegas 4; Socorro Co.: Mag- dalena Mountains ?. Norru Carozra. Ashe Co.: New River 2; Carteret Co.: Boque @ 6; Hyde Co. 3964; Pasquotank Co.: Elizabeth City.* OxtaHoma. Alfalfa Co.: Cherokee ¢?; Carter Co.: Ardmore 3¢; Garvin Co.: Paul’s Valley 39; Mc Curtain Co.: Idabel 2. PrNNsytvania. Allegheny Co. 9. SourH Caro- Lina. Greenville Co.: Greenville ¢. SourH Dakota. Bookings Co.: Bookings ¢. Texas. Aransas Co. 346; Bastrop Co. ¢: Smithville ¢; Bee Co.: Beeville ¢°; Bexar Co. 954: Fort Sam Houston ¢, San Antonio 8°68; Blanco Co.: Johnson City 36, Round Mountain ?; Bosque Co.;* Brazos Co.: College Station 39; Cameron Co.: Port Isabel ¢; Camp Co.: Pittsburg 3 9¢; Collin Co.; McKinney ?; Coryell Co.: Gatesville 29; Dallas Co.: Dallas 29174; Denton Co.: Denton 39; Donley Co.: Claren- don ¢@; Duval Co:: San Diego ¢; Eastland Co. 2: Frio Co. 6°44; Galveston Co.: High Island 22; Gillespie Co.: Stonewall 4; Goliad Co.: Goliad ¢; Gonzales Co.: Palmetto State Park 82 6; Grayson Co.: Sherman ¢?; Harris Co.: Baytown @ ; Hemp- hill Co.: Canadian ¢; Henderson Co.: Athens °; Hidalgo Co.: Alamo ¢; Hunt Co.: Quinlan ¢?; Jefferson Co.: Sabine (10 mi. S.) @; Kaufman Co.: Rosser 39; Kerr Co.: Kerrville 29764; Kleberg Co.: Kingsville 89, Los Olmos Creek ¢; Lamar Co.: Baris 426." lea Salle "Co; + Cotulla. 36" Lavaca Cox) Sweet Home ¢; Lee Co. 2346: Fedor 92346, Giddings 289276, Old Dime Box 346; Limestone Co.: Mexia ?; Matagorda Co.: Bay City ¢, Matagorda 29; McLennan Co.: Eddy ¢ 6, Mc- Gregor 26, Waco 2?; Medina Co. ?; Mills Co.: Goldthwaite 2; Montague Co.: Forestburg 2°24; Montgomery Co.: Mag- nolia 4?; Nueces Co. 26: Corpus Christi @, Corpus Christi (Naval Air Station) @; Red River Co.; Clarksville 9; Refugio Co.: Tivoli (5 mi. S.W.) 44¢@ 6; Tarrant Co.: Fort Worth 9°; Travis Co.: Austin 339134; Trinity Co.: Trinity ¢; Uvalde Co.: Concan ¢, Sabinal @; Val Verde Co.: Devil’s River 2¢; Victoria Co.: Victoria @; Webb Co.: Laredo ¢64; Wichita Co.: Wichita Falls ¢; Williamson Co.: Round Rock ¢?, Taylor g. Vircinta. Norfolk Co.: Dismal Swamp (Stumpy Lake?) 1?. Mexico. Coahuila: Piedras Negras (192km.S.) °. FLower Recorps. ANACARDIACEAE: Rhus, sp. 3264, R. copal- lina @. CLETHRACEAE: Clethra sp. 9° 6. ComposiTaE: Aster 316 THe UNIVERSITY SCIENCE BULLETIN sp. 2, Cirsium sp. 72 6, C. altissimum ¢?, C. vulgare,* Coreopsis cardaminefolia 968, Engelmannia pinnatifida 4, Gaillardia sp. 129204, Grindelia sp. 2, G. squarrosa,* Helenium autumnale 39, Helianthus sp. 2 6, H. Maximiliani,* Ratibida columnifera ¢ , R. pinnata ¢, Solidago sp. 139462, S. rigida,* yellow Compositae 6. CONVOLVULACEAE: Euipomoea sp. 9. EUPHORBIACEAE: Eu- phorbia marginata °. HypropHyLLAcEAE: Phacelia sp. 4492¢. LasiaTaE: Brazoria truncata 2, Marrubium vulgare 2? ¢, Mentha longifolia,* Monarda sp. ¢, M. citriodora 22, M. punctata 364, Pycnanthemum virginianum,* P. flexuosum.* | LEGUMINOSAE: Amorpha fruticosa 2, Cassia cinerea 22, C. fasciculata 2°, Dalea aurea 6, D. multiflora 24, Desmodium paniculatum,* Medicago sativa 2276, Melilotus alba 10264, Petalostemum sp. ?, P. purpureum,* P. microphyllum,* Prosopis sp.* LimiackaE: Yucca glauca.* MatyaceaE: Callirhoé leiocarpa é , Gossypium sp.2¢9 ¢. OLEACEAE: Ligustrum sp. 2. ONAGRACEAE: Oenothera biennis.* PotyconaceaE: Polygonum sp. ¢. PONTEDERIACEAE: Pontederia cordata 2. RustacEaE: Borreria sp. 36. SOLANACEAE: Lycium sp. 6, L. halimifolium,* Solanum sp. 2°, S. rostratum 3¢, S. Torreyi 22. UMBELLIFERAE: Eryngium sp. é, E. yuccifolium.* VERBENACEAE: Verbena officinalis 4°. As shown above, N. n. nortoni has been found on a large variety of flowers. These data are summarized in table 3. Cockerell (1934b) states that Parks observed this subspecies on 83 species of plants in Texas. Parasites. Only one parasite has been recorded from this sub- species, the mite Disparipes texanus (Cockerell, 1910a, p. 276), found on the adult bee. Dr. E. A. Cross identified mites from the mesothoracic hair of a male from Austin, Texas, as Imparipes americanus (Acarina: Scutacaridae). Mr. R. C. Funk identified mites from the vertex of the head of a female from Athens, Texas, as Glyphanoetus sp. (Acarina: Anoetidae). Brotocy. Some 20 years ago the late Dr. R. H. Beamer collected two cell clusters of N. n. nortoni near Lawrence, Kansas. Dr. C. D. Michener has kindly turned these cell clusters over to me for dis- section and study. The cell clusters were found, along with many others, in a recently plowed corn field; they must have been thrown out by the plow. There must have been a recent rain, because two cells were partially filled with soil that seems to have been washed into them. These plugs of soil could only have been BANDED SUBGENERA OF NOMIA IN AMERICA 317 TaBLE 3.—Summary of Flower Records for Nomia nortoni nortoni Cresson Plant Data* Records of NV. n. norton : Number | Number | Number | Number | Number Total FAMILY of of of col- of of number genera species | lections | females males of bees Anacardiaceae 1 1 2 4 6 10 Clethraceae.... 1 1 1 1 1 2, Compositae.... 10 12 18 4] 35 76 Convolvu- laceae..... 1 1 1 0 I Euphorbiaceae 1 1 1 1 0 1 Hydrophyl- laceae..... 1 1 2 44 2 46 Labiatae...... 5 7 5 6 4 10 Leguminosae... 8 11 13 18 16 34 Liliaceae...... 1 1 0) (0) 0 0: Malvaceae... . Dy 2 3 D 2, 4 Oleaceae...... 1 1 1 1 =| 1 Onagraceae... . 1 1 0 0 0 0) Polygonaceae. . 1 1 1 0 i 1 Pontederiaceae 1 1 1 1 0) 1 Rubiaceae..... 1 1 2 0 3 3 Solanaceae.... 2 3 4 a 1 8 Umbelliferae. .. 1 1 1 0) 1 1 Verbenaceae. . . 1 1 1 4 Ones 4 sO valle 40 48 57 131 72 203 * Records from the literature are included under this heading. formed when the cells were inverted after being thrown out by the plow. Thus the cell clusters have been weathered, and possibly the years of storage have further changed their shape. Since the cell of N. nortoni has never been described, these cell clusters pro- vide some important information on the biology of this species. The cell cluster seems to be of the type that is surrounded by a cavity and supported upon pillars with the cells in vertical orienta- tion, as shown by Sakagami and Michener (1962, fig. 2) for the Asiatic N. punctata Westwood. They mention that another Asiatic species, N. ruficornis Spinola, has a similar type of nest (p. 47). High (1921) states that the irregular cell cluster of N. nortoni is surrounded by a cavity, but Parks (1928) does not mention such a cavity. The larger cell cluster (diagram a) from Lawrence, which is 57 millimeters high by 83 millimeters long by 60 millimeters wide, contained nine cells; two open but at one time closed, two open and never closed, four with pollen balls, and one with pollen and large larva. The smaller cell cluster, which is not complete, con- 318 THe UNIVERSITY SCIENCE BULLETIN tained six cells; three open and partly full of dirt, two with adult females, and one side of another cell. High states the cell clusters contained from four to fourteen cells, the average being six or eight. The cell clusters are made of hard clay and the cells are lined with the same material. Parks mentions that the cells and burrow are lined with “foreign” soil. The walls of the cells are coated with a thin layer of hard, dark material. The cells are vertical, oval in shape, one side being flattener than the others, with a length of 17 to 19 millimeters and a width of 9.5 to 11.0 millimeters (diagram b). The minimum distance between cells was two millimeters, the maximum was much more than this. The completed cells are closed by a plug that is concave on the inside, 6 to 11 millimeters in length, and 6.5 to 7.5 millimeters in diameter at the bottom, becoming wider at the top. The feces are deposited in fine ribbons at the bottom of the cell; this fecal mass becomes very thin as it continues up the sides of the cell. For burrow depth see High (1921), Parks (1928) and Cockerell (1934b). 20mm. (A= DracraM a. Cross section of N. n. nortoni cell cluster. DiacraM b. Cell of N. n. nortoni. N. nortoni nests in dense aggregations and build up large popu- lations at certain times, as does N. melanderi, but unlike melanderi, nortoni does not seem to prefer alkali soil. Parks found them nesting in cotton and corn fields, and High found them in a sweet potato field. Dr. H. V. Weems, Jr. (in litt.) reports N. nortoni nesting by the hundreds in a lawn at Morriston, Florida. Although nortoni cell clusters are surrounded by a cavity, melanderi has no such cavity (Stephen, 1959a; Bohart, 1952). BANDED SUBGENERA OF NoMIA IN AMERICA 319 SEASONAL Activity. Table 5 gives the seasonal activity of the adults. The specimen bearing the earliest date was taken at Dallas, Texas. N.n. nortoni seems to occur about a month earlier in Texas than in any other state. The latest date in table 5 was based on a specimen collected in Texas (no lacality given). The next latest date, October 14, was from a specimen collected at Cache, Illinois. Nomia (Acunomia) nortoni cressoni Westwood, new combination (Fig. 8; Map 4; Table 5) Nomia cressoni Westwood, 1875, p. 218 (tax., 6, ?); Vachal, 1897, p. 92 (im part) (tax., 9); Cockerell, 1898d, p. 33. (tax.); Cockerell, 1899b, p. 7 (cat.). Nomia nortoni; Patton, 1879, p. 366 (in part) (tax.); Dalla Torre, 1896, p. 168 (in part) (tax.); Rau, 1943, p. 645 (rec.). Nomia (Paranomia) nortoni: Cockerell, 1910b, p. 296 (in part) (axe) Nomia (Acunomia) nortoni: Michener, 1951, p. 1128 (in part) (cat. ). The status of this form has been confused, largely due to Cresson’s description of nortoni, which is discussed above. The characters that separate this form from n. nortoni have not been previously recognized, but are important enough to give this form subspecific status, as seen by the description below. Mate. (8) Tegula slightly emarginate or usually straight on outer edge. (12) Hind tibial lobe with raised area of outer surface extending towards tip, raised area forming distinct ridge; lobe less than 1.7 mm. wide (fig. 8). (19) Hairs long, longest hairs of face about two-thirds as long as scape. (20) Anterior margin of genal area, along eye, with short, appressed, plumose hair. (26) Longest hairs of tergum two twice as long as lateral width of colored bands. FeMALE. (37) Scutellum slightly depressed in center. (40) Tegula straight on outer edge. (45) Hairs long, longest hairs of face longer than distance from antenna to inner eye margin. (46) Anterior margin of genal area, along eye, with short, appressed, plumose hair. (48) Median mesoscutal line obscured anteriorly by long hairs. VariATION. This subspecies exhibits little interpopulational vari- ation in Mexico. It may intergrade with n. nortoni in northeastern Mexico or in north central Mexico. Evidence for the latter possi- bility is the males from southern Arizona that have short body hair like n. nortoni and an intermediate hair condition on the genal area, 320 THe UNIVERSITY SCIENCE BULLETIN but have the hind tibia as in n. cressoni. The females from Arizona also have body hair like n. nortoni, but have a genal hair condition like cressoni. Type MatertaLt. Westwood in his original description cites a male from Mexico (D. Coffin) labeled Hym. No. 96. He also mentions the female. The male is in the Oxford Museum collec- tion together with a female labeled Mexico (D. Coffin). This female specimen seems to be a syntype. It is not known whether Westwood also had other specimens when he described cressoni. I now designate this male as the lectotype for cressoni. Dr. M. W. R. de V. Graham compared this specimen with my description of n. cressoni and with a specimen from San Juan de los Lagos and reports agreement. Rance. This subspecies is found on the Mexican plateau, above 5,000 feet, from Mexico City to Durango, with possible intergrades in southern Arizona (map 4). List oF Locauitizs. ARIZONA (intergrades toward n. nortoni; see discussion above). Santa Cruz Co.: Madera Canyon (Santa Rita Mountains) ¢, Patagonia 2¢, Pema Blanca Lake (5.5 mi. W.) 62, Ruby 2926, Sycamore Canyon (near Ruby) 72136. México. Distrito Federal 2¢ ¢: Guadalupe ¢, Mexico City 3944; Durango: Coyotes ¢, El Salto (6 mi. N. Eo) 13 Ralos Colorados ¢; Guanajuato: Jural ¢; Jalisco: Guadalajara 2?2¢, Lagos de Moreno (15 mi. N.E.) 6, La Piedad ¢, Ojuelos ¢, San Juan de los Lagos 52134, Villa Guadalupe ¢2¢; Mexico: Texcoco; * Michoacan: Lake Paétzcuaro 92; Puebla: Puebla (28 mi. W.) ?; Zacatecas: Fresnillo (9 mi.S.) ¢. FLOWER Recorps. ASCLEPIADACEAE: Asclepias sp. 9246. Lasta- TAE: Stachys Bigelovii ¢°. LrcumiINosaE: Eysenhardtia poly- Stachys 2956. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Acunomia ) melanderi Cockerell This species of Acunomia is similar to nortoni, but is much smaller. It can be distinguished from any other North American bee, except robinsoni, by its size, integumental colored abdominal bands, long pointed antennae in the male and the coarse irregular teeth near the apex of the middle tibial spur in the female. From robinsoni it can be told by the lack of a carina on the pronotum. BANDED SUBGENERA OF NOMIA IN AMERICA 321 Mate. Structure: (1) Mandible with large toothlike lobe on posterior margin. (2) Clypeus with bare part impunctate or with few fine, scattered punctures; bare part flat or slightly depressed in center or with weak longitudinal carina; subapical area slightly depressed or flat. (3) Last flagellar segment acutely pointed, one and one-half times as long as second flagellar segment. (4) Prono- tum without carina on posterior, dorsal edge. (5) Scutellum slightly depressed in center, lateral higher parts less densely punc- tate than surrounding area. (6) Metepisternum coarsely to finely wrinkled. (7) Propodeum with anterior groove slightly widened medially. (8) Tegula straight on outer edge, rounded posteriorly. (9) Front wing length (see section on methods) about 8.0 mm. (range: 7.0-9.6). (10) Middle tibial spur straight. (11) Posterior femur twice as long as broad, large basal and lateral areas nearly impunctate and shining. (12) Posterior tibia produced anteriorly and distally into large rounded lobe which hides tibial spurs in side view (fig. 9). (13) Metasomal tergum 1 coarsely punctate except for band of minute punctures at base of apical, depressed, hyaline portion; median width of band varies from two to four rows of punctures, width of band reduced laterally. (14) Terga 2 to 4 with few coarse punctures like those of tergum 1 on raised part of postgradular area; otherwise with small, close punctures except on colored bands. (15) Sternum 3 with ventrolateral, tooth- like process three to four times as wide as high. (16) Ventroapical median part of gonobase forming rounded projection. (17) Penis valve with small oblong ventral opening; valve produced laterally into rounded projection; tip slender and pointed (fig. 31). Pubes- cence: (18) Face, including upper half of clypeus and lateral part of scape, covered with short, appressed, plumose, white hairs. (19) Longest hairs of face half to two-third as long as scape. (20) Anterior margin of genal area, along eye, with short, ap- pressed, plumose hairs. (21) Hair of anterior part of mesoscutum, of posterior lobe of pronotum, and of anterior part of mesepi- sternum yellow-white to white. (22) Median mesoscutal line and notaulus obscured by hairs or visible posteriorly. (23) Metanotum with anterior fourth to half obscured by short, plumose hairs. (24) Tegula covered with plumose hairs posteriorly. (25) Pro- podeal hairs white. (26) Longest hairs of tergum 2 as long as to twice as long as lateral width of colored band. (27) Terga 2 to 4 with hairs all white. Color; (28) Mandible black or translucent By THe UNIVERSITY SCIENCE BULLETIN red-brown, with subapical red area. (29) Flagellum yellow to orange below. (30) Legs black to red-brown, except metabasi- tarsus below, other tarsal segments, and underside and apices of tibiae, which may be yellow. (31) Stigma usually dark brown, darker than most wing veins; vein R, intersection of M and Cu, and distal end of V dark brown to black. (82) Terga 2 to 5 and sometimes 1 with apical colored bands; band of tergum 5 present laterally; band color variable, blue, green, white or yellow with pink, gold, purple or red iridescence. FEMALE. Structure: (33) Clypeal margin truncate to emar- ginate medially; part overhanging base of labrum hyaline or not; longitudinal carina very weak; clypeus evenly punctate or punctures weaker and further apart on central elevated portion. (34) Face with elevated portion below antennae impunctate or with few punctures. (35) Pronotum without carina on posterior, dorsal edge. (36) Mesoscutum and scutellum with large bare nearly impunctate areas; scutellum flat in center. (37) Mesepisternum with medium to coarse wrinkles. (38) Propodeum with anterior groove sometimes widened medially. (39) Tegula straight on outer edge, rounded posteriorly. (40) Front wing length (see section on methods) about 7.2 mm. (range 6.2-8.4). (41) Metasomal tergum | with small punctures at sides, becoming smaller and more scattered towards center except band of minute punctures at base of apical, depressed, hyaline portion; median width of band half width of hyaline portion. (42) Tergum 2 with small punctures at sides of basal area, becoming much finer toward center or not be- coming finer toward center. (43) Tergum 3 with gradular ridge distinct and with few scattered punctures. Pubescence: (44) Longest hairs of face about as long as distance from antenna to inner eye margin. (45) Anterior margin of genal area, along eye, with short, appressed, plumose hairs. (46) Mesoscutal hairs all white to mostly dark. (47) Notaulus visible or obscured; median mesoscutal line visible or obscured anteriorly by long hairs. (48) Metanotum completely to its anterior half obscured by short, plu- mose hairs; no longer black hairs present. (49) Tegula covered with plumose hair posteriorly. (50) Tergum 1 with lateral part of depressed, hyaline, portion partly obscured by subappressed mat of hairs. (51) Terga 2 and 3 with hair all white, tergum 4 with or without some black hairs. Color; (52) Antennae light brown to orange-brown below. (53) Stigma black, darker than most veins; vein R, intersection of M and Cu, and distal end of BANDED SUBGENERA OF NOMIA IN AMERICA 323 V black to dark brown. (54) Legs dark brown. (55) Terga 1 to 5 with apical colored bands, sometimes indistinct on tergum 1; band color variable, yellow, white or green with purple, pink, yellow, orange or red iridescence. Nomia (Acunomia ) melanderi melanderi Cockerell (Figs. 9, 25, 27-29 and 31; Map 4; Tables 4 and 5) Nomia melandri Cockerell, 1906b, p. 279 (lapsus calami) (tax., 2); Tysdal, 1940, p. 583 (alf.); Linsley and MacSwain, 1947b, p. 354 (alf.); Michener, 1954, p. 68 (pupa). Nomia melanderi: Cockerell, 1908, p. 334 (justified emendation of N. melandri Cockerell) (tax., @ ); Cockerell, 1925a, p. 188 (tax.); Cockerell, 1930, p. 12 (tax.); Linsley, 1946, p. 19 (alf.); Vansell and Todd, 1946, p. 473 (alf.); Bohart, 1947, p. 13 (biol.); Linsley and MacSwain, 1947b, p. 352 (alf.); Bohart, 1948, p. 58 (alf.); Annand, 1949, p. 49 (biol.); Bohart and Lieberman, 1949a, p. 519 (alf.); Bohart and Lieberman, 1949b, p. 1 (alf.); Anony- mous, 1950, p. 212 (alf.); Bohart, Knowlton and Bailey, 1950, p. 4 (rec.); Bohart, Knowlton, Nye and Todd, 1950, p. 66 (biol. ); Hoyt, 1950, p. 58 (alf.); Linsley, MacSwain and Smith, 1950, p. 61 (alf.); Bohart, 1951, p. 32 (biol.); Hoyt, 1951, p. 61 (alf.); Menke, 1951, p. 624 (alf.); Bohart, 1952, p. 110 (biol.); Menke, 1952a, p. 16 (biol.); Menke, 1952b, p. 334 (alf.); Anonymous, 1953, p. 21 (biol.); Menke, 1953, p. 15 (alf.); Menke, 1954a, p. 539 (alf.); Menke, 1954b, p. 1 (alf.); Stevens, 1954, p. 33 (alf.); Bohart, 1955b, p. 23 (biol.); Bohart and Cross, 1955, p. 403 (biol. ); Bohart, Nye and Levin, 1955, p. 51 (biol.); Bohart and Nye, 1956a, p. 332 (alf.); Bohart and Nye, 1956b, p. 468 (biol.); Bo- hart, 1957b, p. 359 (alf.); Bohart and Nye, 1957, p. 83 (biol.); Frick, 1957, p. 503 (para.); Todd, 1957, p. 66 (biol.); Bohart, 1958a, p. 94 (biol.); Bohart, 1958b, pp. 931 and 935 (biol. ); Cross, 1958b, p. 135 (biol.); Linsley, 1958, pp. 551-554, 556, 566, 581-584 (biol.); Anonymous, 1959a, p. 150 (para.); Stephen, 1959a, p. 3 (biol.); Bohart, 1960a, p. 75 (alf.); Bohart, 1960b, pp. 59 and 64 (alf.); Bohart and Nye, 1960, p. 12 (fls.); Bohart, Stephen and Eppley, 1960, p. 425 (para.); Frick, Potter and Weaver, 1960, p. 1 (biol.); Fronk and Painter, 1960, p. 424 (biol.); Howell, 1960, p. 679 (biol.); Laffoon, 1960, pp. 176 and 202 (tax.); Stephen, 1960a, p. 15 (biol.); Stephen, 1960b, p. 27 (biol.); Stephen, 1960c, p. 1025 (biol.); Stephen and Evans, 1960, p. 3 (biol.); Frick, 1962, Paon Goalies). 324 THe UNIvERSIty SCIENCE BULLETIN Nomia (Paranomia) melanderi: Cockerell, 1910b, p. 296 (tax.). Nomia (Paranomia) californica Cockerell, 1910b, p. 296 (tax., 2). Nomia (Paranomia) acus Cockerell, 1910b, p. 296 (tax., ¢ ). Nomia acus: Cockerell, 1923, p. 102 (tax.); Cockerell, 1930, p. 12 (tax.); Bohart, Knowlton and Bailey, 1950, p. 4 (rec.). Nomia californica: Cockerell, 1924, p. 541 (tax.); Cockerell, 1925a, p. 188 (im part) (tax.); Blair, 1930; pp:-7 and dij (@axe)e Cockerell, 1930, p. 12 (tax.); Cockerell, 1933, p. 159 (rec.); Cock- erell, 1934a, p. 21 (rec.); Michener, 1936, p. 4 (rec.); Michener, 1937, p. 318 (tax.); Bohart, Knowlton and Bailey, 1950, p. 4 (rec.). Nomia melanderi acus: Cockerell, 1925a, p. 188 (tax.). Nomia melanderi paysoni Cockerell, 1925c, p. 621 (tax., ¢ ); Cockerell, 1928, p. 101 (rec. ). Nomia californica var. paysoni: Cockerell, 1930, p. 12 (tax.). Nomia (Acunomia) melanderi: Bohart, 1950, p. 32 (biol.); Cross, 1958a, p. 1265 (tax.); Cross and Bohart, 1960, p. 761 (biol.); Hirashima, 1961, pp. 294 and 300 (biol. ). Nomia (Acunomia) melandri melandri: Michener, 1951, p. 1128 (cat-). Nomia (Acunomia) melandri: Michener, 1953, p. 1030 (larva). Nomia melanderi melanderi: Selander and Bohart, 1954, p. 228 (biol. ). Nomia (Acunomia) melanderi melanderi: Krombein, 1958, p. 103 (para. ). Alkali bee, Aicher, 1917, p. 5 (alf.); Knowles, 1943, p. 29 (alf.); Tysdal, 1946, p. 531 (alf.); Vanseil and Todd, 1947, p. 137 (alf.); Bohart, 1952, p. 1 (biol.); Hoyt, 1952, p. 49 (alf.); Knowlton, 1952, p: 2 (alf.); Vansell and Griggs, 1952, p,. 90° (ilss)sailonz 1954, p. 418 (alf.); Bohart, 1955a, p. 26 (alf.); Bohart and Nye, 1956c, p. 605, (biol).; Bohart, 1957a, p. 33 (alf.); Robb, 1957, p. 8 (alf.); Wheeler and Hill, 1957, p. 333 (alf.); Anonymous, 1958, p. 8 (biol.); Anonymous, 1959b, p. 12 (biol.); Nye and Bohart, 1959, p. 140 (para.); Stephen, 1959b, p. 69 (biol. ); Anony- mous, 1960a, p. 38 (alf.); Anonymous, 1960b, p. 14 (biol.); Bo- hart and Knowlton, 1960, p. 1 (biol.); Doull and Purdie, 1960, photos on p. 17 (biol.); Morse, 1960, p. 7 (biol.); Stephen, 1960d, p. 30 (biol.); Anonymous, 1961, p. 4 (biol.); Bohart and Todd, 1961, p. 242 (alf.); Stephen and Torchio, 1961, p. 89 (biol.); Todd and McGregor, 1961, pp. 249 and 368 (alf.). BANDED SUBGENERA OF NOMIA IN AMERICA 325 This subspecies is commonly referred to as the alkali bee (Laf- foon, 1960); therefore references using only this name are cited above under that name. The alkali bee has been introduced by means of artificial bee “beds” into several areas where it does not occur or where it occurs in small numbers. These introductions will present prob- lems for future students studying the geographical variation of this bee. In Cockerel'’s original description the spelling melandri is used for this species. The label attached to the holotype also has this spelling. Cockerell stated he received the holotype from A. L. Melander from which one can judge that his spelling of the spe- cific name was a lapsus. Thereafter Cockerell himself used the spelling melanderi; his 1908 publication can be considered as a justified emendation of the name melandri. This subspecies has gone under several other names, as shown above, due in part to the great differences between the sexes and its wide range. Mate. (9) Front wing length (see section on methods) about 8.5 mm. (range: 7.5-10.5). (12) Hind tibial lobe usually squared on anterodistal margin (fig. 9). (19) Hairs long, longest hairs of face about two-thirds as long as scape. (22) Median mesoscutal line and notaulus obscured by hairs. (26) Longest hairs of tergum 2 about twice as long as lateral width of colored band. (32) Colored band of tergum 1 usually absent, sometimes present at sides or complete but narrowed in center, or occasionally broad as in m. howardi. Femate. (42) Tergum 2 with small punctures at sides of basal area, punctures becoming finer towards center. (46) Mesoscutal hairs mostly dark, sometimes all white. (47) Hairs long, notaulus obscured, median mesoscutal lines obscured anteriorly by hairs. (55) Tergum 1 with apical colored band, sometimes indistinct; band one-third width of band of tergum 2; band on tergum 2 0.66 mm. wide. Variation. Like N. n. nortoni, this wide ranging subspecies ex- hibits little interpopulational variation in the characters studied. A specimen collected in Idaho could not be distinguished from one collected in southern California. However, this subspecies exhibits much more intrapopulational variation than n. nortoni. All the characters that separate m. howardi from m. melanderi are occasion- ally found among individuals of m. melanderi, but individuals with 326 THe UNIVERSITY SCIENCE BULLETIN these characters occur at random throughout the range of m. me- landeri, and single specimens do not show more than one or possibly two characters of the other subspecies. This variation does not seem to be clinal in nature. Some of the specimens from the Hemet-La Quinta area (a and b respectively on map 4) in southern California are an exception to the above. Those from Hemet agree with m. melanderi in most characters, but tend toward m. howardi in hair pattern.! The female from La Quinta mostly closely resembles m. melanderi, except in hair length, in which it agrees more closely with m. howardi. The male from the same locality is a typical m. howardi. La Quinta is 38 miles from Hemet. Type MatertaL. The female holotype of melanderi, N. Yakima, Washington, July 10, 1903, Yakima Expedition (from A. L. Me- lander ), is in the collection of the Citrus Experiment Station, River- side, California. The female holotype of californica, Los Angeles Co., California (D. W. Coquillett), is at the U. S. National Mu- seum, type number 12986. The male holotype of acus, southern California (C. V. Riley), is also in the U. S. National Museum, type number 12987. The male holotype of melanderi paysoni, Naturita, Colorado, June 27, 1924 (Edwin Payson), is at the Citrus Experi- ment Station. Rance. This subspecies ranges from central Washington to southern California and east to Wyoming, where it is replaced by the similar but larger eastern species, N. nortoni. To the south m. melanderi is replaced by the desert form, m. howardi (map 4). I have seen four specimens of m. melanderi labeled Sedan, Minne- sota. If these specimens are labeled correctly it seems strange that no specimens of this subspecies have been collected in South Dakota. List oF Locauitres. CatirorntA. Colusa Co.: Colusa 6 2 22 4 Contra Costa Co.: Antioch 1499 ¢; Fresno Co. 928: Dos Palos (5 mi. S.) 9, (5 mi. FE.) 43, (7 mi. E.) 2146, Firebaugh e773, (1 mi. S.) 32, (2 mi. W.) 2, (4mi. N. W.) 36; (7 miss 2 ore Kerman (8 mi. W.) ¢, Mendota 2964, Oxalis 29, Rasin 29, Tranquillity 23; Glenn Co.: Artois 9; Inyo Co.: Bigpine 29 6, Bishop ¢ 6, Lone Pine (2 mi. N.) ¢, (3 mi. N.) 2? 6; Kern Co.: McFarland ¢, Shafter 9; Kings Co.: Lemoore (8 mi. S.W.) 9, 1. Six other females in the University of California collection are typical m. howardi although, labeled as from Hemet. As all were taken on a single day, and as collections of howardi were being made at about the same time by the same collector in Blythe it seems probable that the specimens of howardi from Hemet were mislabeled. If not, the two forms exist together at Hemet and a case might be made for regarding them as distinct species. BANDED SUBGENERA OF NOMIA IN AMERICA Sey Stratford (4 mi. N. W.) 59; Lassen Co.: Hallelujah Junction 9@, Wendel ¢; Los Angeles Co. 22; Merced Co.: Dos Palos 16 9 224, El Nido (4 mi. S.E.) ¢, Los Banos ¢, Merced ?, Turlock (10°mi S: W.) @;-Napa Co:: St. Helena ; 9 6; Riverside Co:: Hemet 752744, (10 mi. W.) 42, Riverside 68 9 40 6, Riverside (Fairmont Lake?) 84, Romoland ¢; Sacramento Co.: Sacra- mento 74, Sherman Island @; San Joaquin Co.: Tracy 39206, (6 mi. N.W.) 9°24; Santa Clara Co.: Milpitas (Dixon Road) 85 2604; Siskiyou Co.: Gazelle 292¢, Lower Klamath Lake 6; Solano Co.: Rio Vista ?; Stanislaus Co.: Newman ¢, Pat- terson é, San Joaquin River (10 mi. S.W. Turlock) ¢, Turlock 152404, Westley 2? ¢; Tulare Co.: Angiola ?; Yolo Co.: Davis 2346, Winters (Putah Creek) ¢, Woodland 2?2¢; Yuba Co.: Marysville 976; County ?: San Francisco Bay (Salt Marshes ) 5292¢. Cororapo. Delta Co.: Delta 59176, Hotchkiss ¢; Eagle Co.: Gypsum 74; Garfield Co.: Glenwood Springs 2°, Rifle 69346; Mesa Co.: Fruita 59 ¢, Grand Junction 29292, Palisade 4924; Montrose Co.: Naturita ¢. Ipano. Ada Co:: Boise (7 mi. S. E.) 52 6; Bear Lake Co.: Paris 9; Bingham Co.: Aberdeen 13986, (5 mi. N.) 119126, (4 mi. E.) ¢, Fort Hall 926, Mill Creek (near Blackfoot) 2°, Springfield ¢ ¢, Sterling (3 mi. W.) 82208; Butte Co.: Howe (11 mi. N.) 4¢; Canyon Co.: Melba (8 mi. N.) 2, (2 mi. N.) 9 6, Parma 24; Franklin Co.: Preston 16¢; Gooding Co.: Bliss 79 ¢, Tuttle 924, Wen- dell ¢; Jerome Co.: Eden 292¢, Hazelton (5 mi. S.) 9, (8 mi. N. E.) 202, (4 mi. N.) 39, (7 mi. N.) 32 6, Jerome 7934, (7 mi. S.) ?; Latah Co.: Moscow ?; Madison Co.: Rexburg 7?; Owyhee Co.: Grand View 2?, Marsing (3 mi. S.W.) @, Murphy? Hot Springs 246; Payette Co.: Fruitland ¢, Payette 44; Power Co.: American Falls @; Twin Falls Co.: Buhl 29 6, Castleford ¢, Clover 26954, Filer ¢, Hansen ¢, Hollister 2 ¢, Kimberly 15954, Twin Falls 12986; County ?: Blue Lakes (Snake River Canyon) 24. Mrnnesota. Pope Co.: Sedan 23 6¢. Montana. Sanders Co.: Weeksville ¢. Nevapa. Humboldt Co.: Paradise Valley ¢; Lincoln Co.: Panaca 29°; Washoe Co.: Pyra- mid Lake 9994. Orrcon. Deschutes Co.: Bend,* Redmond 254; Gilliam Co.: Arlington 2924; Jefferson Co.: Madras 5934; Klamath Co.: Henley 9; Malheur Co.: Adrian 292, (8 TMS) Gey (Gumi Ss) Le oe Stamni. NOW). 226, ajuntura, G3 mi. S.) ¢, Nyssa (Klamath Ave.) 29, (10 mi. S.W.) ¢, Ontario 6, Vale 928, (West of) 29; Umatilla Co.: Hermiston ¢, 328 THe UNIveRSIry SCIENCE BULLETIN Milton,* Umatilla 54; Union Co.: Hot Lake 3?96¢, La Grande 2°, North Powder 8936; Wasco Co.: The Dalles.* Urau. Beaver Co. 924; Box Elder Co.: Brigham City ¢, Garland ¢, Tremonton (10 mi. S.W.) 4224; Cache Co.: Ballard Junction 926, Logan 199138, Mendon ¢, Newton @ 4, Petersboro,* Smithfield 8@ ¢; Carbon Co.: Pleasant Valley (near Scofield) 2 6, Price 5968; Duchesne Co.: Bluebell 7? ¢, Myton 18 ¢ 44, Roosevelt 934; Emery Co. 9326: Castle Dale 29242; Iron Co.: Iron Springs ¢ 6, Little Salt Lake ¢, Parowan 5¢; Juab Co.: Trout Creek ¢; Kane Co.: Orderville 24, Esclante Desert (Willow Tank?) ¢; Millard Co. 2?: Delta 33961 ¢, Fillmore 2¢, Flowell 24, Gandy 3?, Leamington 4¢, Lynndyl 2.6, (sand dunes N.)° 952, Pahvant. 263, Sutherlandiatoear Topaz 244; Piute Co.: Circleville 3224; Rich Co.: Bear Lake Valley (west side) 79; Salt Lake Co.: Salt Lake City 2924; Tooele Co.: Iosepa °, St. John ¢; Uintah Co.: Jensen ¢, Rand- lett 54. Vernal 59. Utah/Co:: Provo 3'9 2'4. Thistles2;oeaUtaln Lake (west side) ¢, (east side) ¢; Washington Co.: Pintura 9, St. George 39 6; Weber Co.: Ogden ¢. Wasuincton. Adams Co.: Othello; * Benton Co. ¢@: Prosser 2? 6; Franklin Co.: Scootenary Reservoir; * Grant Co.: Moses Lake 3?, Quincy 9°, Soap Lake; * Walla Walla Co.: Gardena 5246; Whitman Co.: Pullman ¢; Yakima Co.: Harrah 39, Wapato 2¢, Yakima 2924. Wyominc. Big Horn Co. (north) 4234: Greybull ¢; Campbell Co. ¢; Fremont Co.: Lander (15 mi. S.) ¢@, Pavillion 39, River- ton 16¢ ¢, Shoshoni 83?4¢4; Park Co.: Powell 29364; Sweet- water Co.: Granger 6924, Green River 934; Teton Co.: Grand Teton National Park 9@3¢4; Yellowstone National Park 39°3¢: (Dunraven Pass) 6, (Tower Falls) @ 6, (Old Faithful) ¢. FLower Recorps. ASCLEPIADACEAE: Asclepias sp. 5956, A. speciosa 6, A. subverticillata 9. BoracrtnaceaE: Helitropium sp. 234, H. curassavicum 42 6. CAPPARACEAE: Cleome sp. 32 4¢,C. lutea 2 6, C. serrulata 149294. CHENOPODIACEAE: Beta vulgaris 32108, Salsola Kali ¢. Composirar: Aster sp. 79, Baccharis emoryi 22, Centromadia pungens 2°54, Chrysotham- nus, 2944, C. nauseosus,” Cirsium sp. 22, Coreopsis sp. 2, Crepis glauca,* Erigeron sp. 4, Grindelia sp. 39°, Haplopappus sp. @, H. venetus ¢, Helianthus sp. 2946, Isocoma veneta ¢, Pluchea camphorata 6254, P. sericea 79, Solidago sp. 3296, S. occidentalis 22. CONVOLVULACEAE: Convolvulus arvensis,* Cressa cretica 6, Cuscuta sp.,* C. salina.* CRructIFERAE: Brassica sp. 32, B. geniculata 9, Sisymbrium altissimum @, Wislizenia BANDED SUBGENERA OF NOMIA IN AMERICA 329 refracta 3°. CucuRBITACEAE: Citrullus vulgaris 2936. Dipsaca- CEAE: Dipsacus.* EUPHORBIACEAE: Croton californicus 4 6 , Eremo- carpus setigerus 2. LapiaTaE: Mentha sp. 9, M. spicata.* LrecuMINOSAE: Astragalus sp. ?, Lotus sp. 3936, L. purshianus 9144, Medicago sp. °, M. sativa 562224, Melilotus sp. 179 406, M. alba6°266, M. indica,* Trifolium sp. 6, T. hybridum ¢, T. pratensa 9, T. repens @. LiiacearE: Allium sp.* Matva- CEAE: Gossypium sp. 9. PAPAVERACEAE: Eschscholtzia californica 29. PoLyconaceaE: Eriogonum sp. 132 6, Polygonum lapathi- folium 2134. SotanaceaE: Solanum tuberosum 9° é, S. elae- agnifolium 92. TAMARICACEAE: Tamarix gallica 2248. Umn- BELLIFERAE: Apium graveolens ¢, Berula erecta & , Daucus carota 59 6. VERBENACEAE: Lippia lanceolata 52. ZYGOPHYLLACEAE: Tribulus sp. 28. As shown above, m. melanderi is found on a wide variety of flowers. Bohart and Nye (1956c) say they have found as many as six different types of pollen on the legs of a single bee. For more information on floral preferences see Selandar and Bohart (1954) and Stephen (1959a). The data above are summarized in table 4. PARASITES, PREDATORS AND DisEAsEs. In the list below the life stage that is attached is shown in brackets. Only the first citation in the literature is given. Bacteria [larva and prepupa], Stephen, 1959a. Aspergillus ustus (fungus ) [prepupa], Stephen, 1959a. A. terreus (fungus ) [prepupa], Stephen, 1959a. Imparipes americanus (Acarina: Scutacaridae) [adult], deter- mined by Dr. E. A. Cross. Mites [larva and prepupa], Stephen, 1959a. Cicindela haemorrhagica (Coleoptera: Cicindelidae) [adult], Frick, 1957. C. pusilla (Coleoptrea: Cicindelidae ) [adult], Frick, 1957. Nemognatha sp. (Coleoptera: Meloidae) [egg and pollen ball], Bohart, 1950. Zonitis atripennis (Coleoptera: Meloidae) [egg and pollen ball], Selander and Bohart, 1954. Hetrostylum robustum (Diptera: Bombyliidae) [larva], Bohart, 1950. Erax stramineus (Diptera: Asilidae) [adult], Frick, 1962. Phoridae (Diptera ) [egg and pollen ball], Bohart, 1950. Zodiom obliquefasciatum (Diptera: Conopidae) [adult], Bo- hart, 1950. Otitidae (Diptera) [pollen ball], Bohart, 1950. 330 Tue UNIVERSITY SCIENCE BULLETIN TABLE 4.—Summary of Flower Records for Nomia melanderi melanderi Cockerell Plant Data* Records of N. m. melanderi Number | Number | Number | Number | Number Total FAMILY of of of col- of of number genera species | lections | females males of bees Asclepiadaceae I 2 7 6 6 12 Boraginaceae. . 1 1 6 5 4 9 Capparidaceae 1 2 10 18 34 52 Chenopodi- aceae...... 2, 2 11 3 11 14 Compositae.... 14 15 36 47 32 79 Convolvu- laceae..... 3 3 1 0 1 1 Cruciferae... .. 3 3 7 8 0 8 Cucurbitaceae 1 1 2 2 3 5 Dipsacaceae. . . 1 1 0 0 0 0 Euphorbiaceae 2 2 2 1 4 5 Labiatae..:... 1 1 1 1 0 1 Leguminosae. . 6 8 60 88 86 174 Liliaceae...... 1 1 0 0 0 0 Malvaceae... . 1 1 1 1 0 1 Papaveraceae. . 1 1 2 2 0 2 Polygonaceae. . 2 2 9 15 14 29 Solanaceae... . 1 2 3 2 1 3 Tamaricaceae. . 1 1 2 2 4 6 Umbelliferae. .. 3 3 6 5 3 8 Verbenaceae. . . 1 1 2 5 0 5 Zy gophyl- laceae..... 1 1 1 0 2 2 AOU Se 48 54 169 211 205 416 * Records from the literature are included under this heading. Sarcophagidae ( Diptera ) [pollen ball], Bohart, 1950. Dasymutilla vesta (Hymenoptera: Mutillidae) [?], Krombein, 1958. Pseudomethoca propinqua (Hymenoptera: Mutillidae) [?], Krombein, 1958. Formica fusca (Hymenoptera: Formicidae) [adult], Frick, 1962. Nomada suavis (Hymenoptera: Apidae) [pollen ball], Bohart, 1950. Skunk (Carnivora: 1959a; robin, Frick, 1962; lark, Frick, 1962. Mustelidae) [prepupa], Anonymous, 1958. Mus musculus (Rodentia: Muridae) [prepupa], Frick, 1962. Microtus sp. (Rodentia: Cricetidae) [prepupa], Frick, 1962. Birds [adult]: blackbirds, Anonymous, 1953; English sparrow, Stephen, 1959a; magpie, Stephen, 1959a; meadowlark, Stephen, BANDED SUBGENERA OF NOMIA IN AMERICA 331 Dr. G. E. Bohart has kindly supplied the following unpublished biological information on Euphytomima nomiivora James (1955) (Diptera: Sarcophagidae), a parasite of Nomia melanderi, Noma- dopsis anthidius and N. scutellaris. It is found in northern Utah and extreme southern Idaho, where it seems to be a serious parasite of melanderi. It lays eggs and first-instar larvae in the lip of the burrow. The larvae probably are carried down to the cells by the bee. They rupture the egg of the bee and then feed on the pollen ball. The mature maggots migrate several inches away from the bee cell to pupate. Brotocy. Since this subspecies is an important pollinator of alfalfa much has been written about its biology, the most important works being by Bohart, Stephen, Frick and Menke. SEASONAL Activiry. Table 5 gives the seasonal distribution of collection dates on the specimens examined. This range consider- TABLE 5.—Seasonal Activity and Number of Specimens Examined of Acunomia and Curvinomia Earliest date Latest date Total collected collected number SPECIES = of bees Month Day Month Day examined N.maneet.....:... V 14 IX 1 105 N. fedorensis....... V 11 VII 17 56 N. universitatis..... V 2 VIII 16 110 N. angustitibialis...| VIII 1 IX 27 62 Nifortt bet Heke VI 15 x 24 898 N. t. tetrazonata... . Ill 25 xX 28 651 N. t. waldensis..... | III 20 VIII 25 119 INE Ws NOntont IV 1 xX 19 743 N.n. cressoni...... VII 0) IX 1 89 N.m. melanderi... . IV 15 XI 6 1,599 N.m. howardi...... IV 6 IX 28 297 N. robinsoni*...... VI 9 X 11 50 EL EXOV eee | se rea ates ete eae ee Wes Oaiecra ml WEN net gesagt ear ee ey eel Ae eas ar 4,779 * Cuba only. ably exceeds the length of seasonal activity given in the literature. Below are some samples of seasonal activity as described by various authors studying the biology of this bee: Washington; late May and mid-June to August (Frick, Potter and Weaver, 1960). Oregon; late June to September 6, 1957 (Stephen, 1959), last week of May to August 26, 1958 (Stephen, 1959). 332 THe UNIvEeRSITY SCIENCE BULLETIN Utah, Idaho and Wyoming; about July 1 for male and middle to late July for females to early September (Bohart, Nye and Levin, 1955). Stephen (1959) and Bohart, Nye and Levin (1955) state that the males emerge slightly before the females, and Stephen shows variation in emergence during different years. Stephen says there is but one brood a year in Oregon, while Bohart, Nye and Levin say there is a partial second generation in Utah. It is not surpris- ing that in examining specimens collected over a number of years in different localities, there is an extension of the dates of adult activity beyond those recorded in a limited area in a short time. The extremes in table 5 are represented by few specimens. The earliest date is from Utah and the latest is from Washington. Nomia (Acunomia ) melanderi howardi Crawford, new combination (Fig. 10; Map 4; Table 5) Nomia howardi Crawford, 1911, p. 617 (tax., 9? ); Cockerell, 1928, p. 103 (rec. ); Cockerell, 1924, p. 541 (tax.); Cockerell, 1930, p. 14 (tax.). Nomia californica: Cockerell, 1923, p. 102 (tax.); Cockerell, 1925a, p. 188 (in part) (tax.). Nomia howardi vanduzeei Cockerell, 1924, p. 540 (tax., @ ); Cockerell, 1930, p. 14 (tax.), new synonymy. Nomia californica peninsularis Cockerell, 1930, p. 12 (tax.), new synonymy. Nomia: peninsularis: Linsley, 1946, p. 20 (biol.); Linsley and MacSwain, 1947a, p. 362 (alf.); Linsley and MacSwain, 1947b, p. 302 (alf.); Bohart and Lieberman, 1949a, p. 519 (alf.); Bohart and Lieberman, 1949b, p. 1 (alf.); Linsley, MacSwain and Smith, 1950, p. 59 (alf. ). Nomia (Acunomia) melanderi peninsularis: Michener, 1951, p. 128" Ceat: Nomia melanderi: Butler, 1955, p. 2 (biol.). This subspecies has gone under several names, as shown above, due in part to the great difference between the sexes, and to Cock- erell’s seeming failure to examine the type of howardi. N. m. howardi is different in several ways from the typical m. melanderi, as shown below, and occupies a different habitat. It seems best to regard howardi as a subspecies of melanderi at this time (see discussion under melanderi proper ). BANDED SUBGENERA OF NOMIA IN AMERICA Soo Mate. (9) Front wing length (see section on methods) 7.5 mm. (range: 7.0-7.9). (12) Hind tibial lobe usually rounded on anterodistal margin (fig. 10). (19) Hairs short, longest hairs of face about half as long as scape. (22) Median mesoscutal line and notaulus not obscured by hairs posteriorly. (26) Longest hairs of tergum 2 about as long as lateral width of colored band. (32) Colored band present on tergum | or band absent (one specimen). FemMae. (42) Tergum 2 with small punctures at sides of basal area, punctures usually not finer towards center. (46) Mesoscutal hairs all white. (47) Hairs short, notaulus and median mesoscutal line visible at least posteriorly. (55) Tergum 1 with apical colored band one-half width of band of tergum 2; band on tergum 2 0.50 mm. wide. VaRIATION. Much the same type as m. melanderi. Scattered individuals of m. howardi show some tendency towards one or two of the characters of m. melanderi, but are recognizable as m. how- ardi by their totality of characters. The characters studied do not vary in clinal fashion; populations near the range of m. melanderi are no nearer to m, melanderi than those from distant localities (for an exception to this see discussion under m. melanderi on the populations in the Hemet-La Quinta area of southern California ). Type Martertat. The female holotype of howardi, Guaymas, Mexico (L. O. Howard), is at the U. S. National Museum, type number 13455. The male holotype of californica peninsularis, Las Animas Bay, Baja California, May 8, 1921 (E. P. Van Duzee), is also there with the type number 55866. The female holotype of howardi vanduzeei, Loreto, Baja Califonia Territorio Sur, May 20, 1921 (E. P. Van Duzee), is in the California Academy of Sciences, type number 1509. Rance. This subspecies is found in the desert or desert scrub of the southwestern United States and northwestern Mexico (map 4). This desert is also called the creosote bush desert after the dominant plant, Larrea tridentata. Lisr oF Locauities. Arizona. Maricopa Co.: Arlington 4, Agua Caliente (Gila River) ¢, Buckeye (12 mi. E.) 44, Good- year 6, Theba ¢; Mohave Co.: Topock ¢; Pima Co.: Marana 8, Sabino Canyon (24 mi. N. E. of Tucson) @, Tucson ¢; Pinal €o.: Eloy (10 mi. S.W.) 58, (12 mi. S. W.) 3¢, (11 mi. S. W.) 254, Maricopa 2 6, Toltec (10 mi. S.) 944; Santa Cruz Co:: Sonoita 9; Yuma Co.: Bill Williams River 2 ¢, Dateland 492, 334 THe UNIVERSITY SCIENCE BULLETIN Roll 16 2 23 6, Wellton 693 6, Yuma 309146. Catirornia. Im- perial Co. 6934: Bard 3¢, Calexico ?, El Centro ?, Experi- ment Farm (near Meloland) ? ¢, Heber (12 mi. E.) ¢, Palo Verde Valley,* Potholes ¢, Westmoreland ¢3¢,; Inyo Co.: Fur- nace Creek ¢; Riverside Co.: Blythe 4791628, (3 mi. W.) @, La Quinta 92 é, Ripley (6 mi. S.) 2924; San Bernardino Co.: Needles ¢; San Diego Co.: Scissors Crossing (6 mi. N. E. of Banner) 11974, San Felipe Creek ¢. Mexico. Baja California: Bahia las Animas 54; Baja California Territorio Sur: Coyote Cove (Bahia Concepcién) 3244, Loreto @; Sonora: Empalme (16 mi. S.E.) 819108, Guaymas ¢?, (14 mi. N.) 9°. FLower Recorps. BoracinaceaE: Helitropium sp. 142, H. curassavicum 24. ComposiTaE: Aster spinosus 2, Isocoma ve- neta 6, Pluchea sericea 6. LrEcumMtINosaE: Medicago sativa 53 9 23 6, Prosopis sp. 2, P. pubescens 32. MatvaceaE: Gossypium sp. 246. SanicacEaE: Salix exigua ?. SOLANACEAE: Solanum elaeagnifolium 69. TAMARICACEAE: Tamarix sp. 9°, T. gallica 64, T. pentandra 24. VERBENACEAE: Lippia lanceolata 5°. Parasites. Mr. R. C. Funk identified mites from the meso- thoracic hair of a female from Blythe, California, as Imparipes americanus (Acarina: Scutacaridae ). Brotocy. Linsley (1946) mentions the nest of m. howardi. Nothing else has been published on the biology of this subspecies. SEASONAL Activity. Table 5 gives the seasonal activity of the adults. Nomia (Acunomia) robinsoni Cresson (Figs. 11, 26 and 32; Map 4; Table 5) Nomia robinsoni Cresson, 1865, p. 174 (tax., 2,6 ); Fox, 1893, p. 134 (tax.); Dalla Torre, 1896, p. 169 (cat.); Cockerell, 1898d, p. 32 (tax.); Cresson, 1916, p. 109 (tax.); Cockerell, 1930, p. 15 (tax.); Krombein, 1953, p. 4 (tax. ). Nomia wickhamii Ashmead, 1896, p. 28 (tax., 4 ); Cockerell, 1898d, p. 32 (tax.); Ashmead, 1900, p. 303 (cat.); Cockerell, 1930. p. 15 (tax. ), new synonymy. Nomia robertsonii: Ashmead, 1900, p. 303 (lapsus calami) (cat.). Nomia (Paranomia) robinsoni; Cockerell, 1910b, p. 297 (tax.). Nomia (Paranomia) wickhamii: Cockerell, 1910b, p. 297 (tax.). Nomia robinsoni wickhamii: Krombein, 1953, p. 4 (tax.). Nomia (Acunomia) robinsoni: Cross, 1958a, p. 1264 (tax.). Nomia (Acunomia) wickhami: Cross, 1958a, p. 1264 (tax. ). BANDED SUBGENERA OF NOMIA IN AMERICA SoD Krombein (1953) states that wickhamii is “at most subspecifically distinct from robinsoni from Cuba, differing only in lacking ferrigi- nous on the hind femora.” Dr. C. D. Michener, who examined the type of wickhamii for me, says it differs from the typical robin- soni in being darker in leg color and having green bands with yellow iridescence (personal communication). These differences fall within the variation of robinsoni from Cuba. It therefore seems best to regard wickhamii as synonymous with robinsoni until more material from the Bahamas is available or some significant differences are found. Mate. Structure: (1) Mandible without lobe on posterior mar- gin. (2) Clypeus with fine, scattered punctures on bare part; bare part raised or with weak longitudinal carina; subapical area flat. (3) Last flagellar segment pointed or rounded, as long as to one and one-third times as long as second flagellar segment. (4) Pronotum with strong carina on posterior, dorsal edge. (5) Scu- tellum deeply depressed in center, lateral higher parts as punctate or less densely punctate than surrounding area. (6) Metepisternum finely wrinkled. (7) Propodeum with anterior groove slightly or distinctly widened medially. (8) Tegula not marginate on outer edge, subtruncate or rounded posteriorly. (9) Front wing length (see section on methods) 7.3 mm., mean (range: 6.8-7.7). (10) Middle tibial spur curved or straight. (11) Posterior femur three times as long as broad, only basal area nearly impunctate and shining. (12) Posterior tibia produced anteriorly and distally into large lobe with deep apical emargination, which does not hide tips of tibial spurs in side view (fig. 11). (13) Metasomal tergum 1 coarsely punctate except for band of minute punctures at base of apical, depressed, hyaline portion; median width of band varies from three to six rows of punctures, width of band not reduced laterally. (14) Terga 2 to 4 without coarse punctures like those of tergum 1 on raised part of postgradular area; all punctures small and close, apical colored bands not punctate. (15) Sternum 3 with ventrolateral, toothlike process four times as wide as high. (16) Ventroapical median part of gonobase forming rounded pro- jection. (17) Penis valve with small oval ventral opening; valve produced laterally into rounded projection; tip broad and rounded (fig. 32). Pubescence: (18) Face, including upper third to half of clypeus covered with short, appressed, plumose, yellowish hairs; lateral part of scape covered with long erect hairs. (19) Longest hairs of face half as long as scape. (20) Anterior margin of genal 336 THe UNIVERSITY SCIENCE BULLETIN area, along eye, with short, appressed, plumose hairs. (21) Hair of anterior part of mesoscutum, of posterior lobe of pronotum, and of anterior part of mesepisternum yellowish. (22) Median mesoscutal line and notaulus obscured by hairs only anteriorly. (23) Meta- notum obscured by short, plumose hairs. (24) Tegula covered with fine simple hairs posteriorly. (25) Propodeal hairs yellowish. (26) Longest hairs of tergum 2 as long as lateral width of colored band. (27) Terga 4, sometimes terga 2 and 3, with some black hairs. Color; (28) Mandible translucent red-brown, with subapical red area. (29) Flagellar segments orange below. (30) Legs dark brown to orange-yellow except tarsi and underside and apices of tibiae, which are yellow to orange. (31) Stigma brown, about same color as most veins; vein R, intersection of M and Cu, and distal end of V brown to dark brown. (32) Terga 2 to 5 with apical colored bands; band of tergum 5 absent laterally; band color variable, green with orange, gold or yellow iridescence. FEMALE. Structure: (33) Clypeal margin truncate medially; part overhanging base of labrum not hyaline; longitudinal carina weak; clypeus evenly punctate. (34) Face with elevated portion below antennae distinctly punctate. (35) Pronotum with strong carina on posterior, dorsal edge. (36) Mesoscutum and scutellum with scattered punctures; scutellum depressed in center. (37) Mesepisternum with fine wrinkles. (38) Propodeum with anterior groove slightly widened medially. (39) Tegula straight on outer edge, rounded posteriorly. (40) Front wing length (see section on methods) 6.7 mm., mean (range: 6.0-7.4). (41) Metasomal tergum 1 with fine punctures at sides becoming small and more scattered towards center except small oval center area with much closer punctures and band of very minute punctures at base of apical, depressed, hyaline portion; median width of band two-thirds width of hyaline portion. (42) Tergum 2 with minute punctures at side of basal area, punctures becoming much finer toward center. (43) Tergum 3 with gradular ridge indistinct and punctate. Pu- bescence: (44) Longest hairs of face shorter than distance from antenna to inner eye margin. (45) Anterior margin of genal area, along eye, with short, appressed, plumose hairs. (46) Mesoscutal hairs yellowish or white anteriorly and darker posteriorly. (47) Notaulus and median mesoscutal line visible at least posteriorly. (48) Metanotum completely obscured by short, plumose hairs; longer black hairs absent. (49) Tegula covered with plumose hairs posteriorly. (50) Tergum 1 with lateral part of hyaline por- BANDED SUBGENERA OF NoMIA IN AMERICA 337 tion obscured by appressed dense mat of hairs. (51) Terga 2 to 4 with some black hairs. Color; (52) Antennae light brown below. (53) Stigma brown, about same color as most veins; vein R and distal end of V dark brown. (54) Legs brown, sometimes hind tibia and tarsus orange. (55) Terga 2 to 5 with apical colored bands; band color green with gold or yellow iridescence. VaRIATION. There are too few specimens of robinsoni from too few localities to draw many conclusions concerning its variation. There is a high intrapopulational variation in integument and band color in the males. The male ultimate antennal segment is highly variable in shape and length. The specimens from the Bahamas (wickhamii) and Mexico fall within the range of variation of the Cuban material. Type MareriaL. WN. robinsoni was described from one female and three males from Cuba. In 1916 Cresson designated a male lectotype, in the Academy of Natural Sciences of Philadelphia, type number 2136. N. wickhamii was described from two males from Eleuthera Island, Bahamas. One of these, which is labeled holo- type, is in the U. S. National Museum with the type number 12990 (Ditzen collector? ). The other specimen must be lost, for I have not been able to find it, and Krombein (1953) says that wickhamii “is known only from the unique male type.” RANGE. Cuba, eastern Mexico(?) and the Bahamas (map 4). List or Locauities. Banamas. Eleuthera Island ¢. Cusa. Camagtiey: Baragua 6, Camagiiey ¢; Habana: Habana 6944, Santiago de las Vegas 3 4; Las Villas: Zaza del Medio ¢ ; Oriente: Cayamas ¢, Santiago de Cuba ¢ ¢; Pinar del Rio: Cabanas ¢, Pinar del Rio @, San Vicente 28, Vinales 9, (7 km. N.) 2182, (14km.N.) 6, (24km. N.) 52; Province?: Soledad ¢. México. Veracruz: Nogales ¢. Parasites. Mr. R. C. Funk identified mites from the sixth abdominal sternum of a male from Vinales (7 km. N.) as Glyphan- oetus sp. (Acarina: Anoetidae ). SEASONAL Activity. Table 5 gives the seasonal activity of the adults from Cuba. The male (only specimen) from Nogales, Veracruz, Mexico, was collected in April, 1956 (N. L. H. Krauss), and was pinned and labeled at the University of Kansas. As the same collector was in Cuba at about the same time, an error in labeling is a possibility. 338 Tue UNIveRsIty SCIENCE BULLETIN BIBLIOGRAPHY AICHER, L. C. LOL. ANNAND, P. 1949. The production of alfalfa seed in southern Idaho, Univ. Idaho Agr. Expt. Sta. Bull., 101: 1-20. N. Report of the Chief of the Bureau of Entomology and Plant Quar- antine 1949, U. S. Dept. Agr., Washington, D. C., pp. 1-63. ANONYMOUS 1950. 1958. 1958. 1959a. 1959b. 1960a. 1960b. 1961. Honey bees and other pollinating insects, Gleanings in Bee Cul- ture, 78(4) :211-213. ‘Bad lands’ bees, Farm Mangt., 2( 10) :21-22. Helping helpful alkali bees, Agricultural Research, 7(2):8-9. Turtox News, 37(6):150 (no title). Bee beds increase alfalfa seed yields, Oregon’s Agricultural Prog- ress, 6(3):12-13. New homes provided for alkali bee, Crops and Soils, 12(8):38. Global trips for alkali bees?, Agricultural Research, 8(8) :14. Leaf-cutter bee good alfalfa pollinator, Oregon’s Agricultural Prog- ress, 8(2):4-5. ; ASHMEAD, WILLIAM H. 1896. 1899. 1900. Notes on the Hymenoptera. Collected by the Bahama Expedition from the State University of Iowa, Bull. Lab. Nat. Sci., State Univ. Iowa, pp. 28-32. Classification of the bees, or the superfamily Apoidea, Trans. Amer. Ent. Soc., 26:49-100. Report upon the aculeate Hymenoptera of the islands of St. Vin- cent and Grenada, with additions to the parasitic Hymenoptera and a list of the described Hymenoptera of the West Indies, Trans. Ent. Soc. London, 1900( 2) : 207-367. BIRKMAN, G. 1899: List of aculeate Hymenoptera, taken at Fedor, Lee County, Tex., Ent. News, 10(8) :244-245. Buiarr, BEULAR Hix 1930. 1935. In Cockerell and Blair, Rocky Mountain bees. I, Amer. Mus. Novitates, 433:7-9, 17-19. The bees of the group Dienuomia, J. New York Ent. Soc., 43(2): 201-215. Bouart, GreorcE E. 1947. 1948. Wild bees in relation to alfalfa pollination, Farm and Home Sci. [Utah Agr. Expt. Sta.], 8(4):13-14. Beneficial insects in relation to alfalfa-seed production in Utah, Rept. 11th Alfalfa Improvement Conf., pp. 57-65 (mimeographed ). The alkali bee, Nomia melanderi Ckll., a native pollinator of alfalfa, Proc. 12th Alfalfa Improvement Conf., pp. 32-35. Alfalfa seed growers of Utah should protect their wild bees, Farm and Home Sci. [Utah Agr. Expt. Sta.], 12(2):32-33, 37. Pollination by native insects, in Insects, Yearbook of Agr., U. S. Dept. Agr., Washington, D. C., pp. 107-121, plate 52. BANDED SUBGENERA OF NOMIA IN AMERICA 339 1955a. Bees make alfalfa seed, What’s New in Crops & Soils, 8(3):12-13, 26. 1955b. Alkali bees vs. drainage, Farm and Home Sci. [Utah Agr. Expt. Sta.], 16(2):23-24, 39-40. 1957a. Pollination as a research field in entomology, Bull. Ent. Soc. Amer., 3(1) :32-83. 1957b. Pollination of alfalfa and red clover, Ann. Rev. Ent., 2:255-380. 1958a. Transfer and establishment of the alkali bee, Rept. 16th Alfalfa Improvement Conf., pp. 94-98. 1958b. Alfalfa pollinators with special reference to species other than honey bees, Proc. 10th International Cong. Ent, [Montreal], 4:929-937. 1960a. Insect pollination of forage legumes, Advances in Agronomy, 12:72-88. 1960b. Insect pollination of forage legumes, Bee World, 41:57-64, 85-97. Bonart, Georce E. and Earte A. Cross 1955. Time relationships in the nest construction and life cycle of the alkali bee, Ann. Ent. Soc. Amer., 48(5) :403-406. Bouart, Georcer E. and F. V. Lirp—ERMAN 1949a. Effect of an experimental field application of DDT dust on Nomia melanderi, J. Econ. Ent., 43(3):519-522 (same paper as 1949b). 1949b. Effect of an experimental field application of DDT dust on Nomia melanderi, Utah Agr. Expt. Sta. Mimeo, Ser., 364:1-6 (same paper as 1949a). Bowuart, Grorcr E. and Grorcr F. KNOWLTON 1952. Wild bees for pollination of the alfalfa seed crop in Utah, Fact Sheet, Utah State Agr. College Extension Service, 8:1-2. 1960. Managing alkali bees for alfalfa pollination, Utah State Univ. Extension Service Leaflet, 78:1-8. Bouart, G. E., G. F. KNow.Ton and R. S. Bary 1950. Some wild bees of Utah, Utah State Agr. College, Agr. Expt. Sta. Mimeo. Ser., 371:1-10. Bouart, G. E., G. F. KNow.ton, W. P. Nye and F. E. Topp 1950. In Growing alfalfa for seed in Utah, Utah Agr. Expt. Sta. Circular, 125:1-72. Bowart, Georce E. and Witii1AM P. Nye 1956a. Kinds of bees, Gleanings in Bee Culture, 84(6):331-333, 337. 1956b. Bees. Their nests and nesting sites, Gleanings in Bee Culture, 84( 8) :468-472. 1956c. Bees. Foraging for nectar and pollen, Gleanings in Bee Culture, 84(10) :602-606, 639. 1957. Bees. The nursery, Gleanings in Bee Culture, 85(2):82-87. 1960. Insect pollinators of carrots in Utah, Utah State Univ., Agr. Expt. Sta. Bull., 419:1-16. Bouwart, Georce E., WituiAM P. Nye and MarsHatyi D. Levin 1955. Section III. Pollination, in Growing alfalfa for seed, Utah State Agr. College, Agr. Expt. Sta. Circular, 135:42-60. 340 Tue UNIVERSITY SCIENCE BULLETIN Bouart, G. E., W. P. SrepHen and R. K. Epriey 1960. The biology of Heterostylum robustum (Diptera: Bombyliidae), a parasite of the alfaki bee, Ann, Ent, Soc. Amer., 53(3):425-435. Bouart, Georce E. and FranK E. Topp 1961. Pollination of seed crops by insects, in Seeds, Yearbook of Agr., Washington, D. C., pp. 240-246, 368. BRIDWELL, J. C. 1899. A list of Kansas Hymenoptera, Trans. Kansas Acad. Sci., 16:203-211. BrIMLEY, C. S. 1938. The insects of North Carolina, N. C. Dept. Agr., Division Ert., Raleigh, North Carolina, pp. 1-560. Butter, G. D., Jr. 1955. Native bees commonly found in alfalfa in Arizona, Univ. Arizona Agr. Expt. Sta., Dept. Ent., pp. 1-3 (mimeographed). COCKERELL, T. D. A. 1893. The entomology of the mid-alpine zone of Custer County, Colorado, Trans. Amer. Ent. Soc., 20:305-370. 1896. The bees of the genus Perdita F. Smith, Proc. Acad. Nat. Sci. Philadelphia, 48:25-107. 1897. Life-zones in New Mexico, New Mexico College Agr. and Mechanic Arts, Agr. Expt. Sta. Bull., 24:1-44. 1898a. The insect visitors of flowers in New Mexico. II., Zool., ser. 4, 2:311-314. 1898b. Tables for the determination of New Mexico bees, Bull. Sci. Lab. Denison Univ., 11(3):41-73 (same paper as 1898c). 1898c. Tables for the determination of New Mexico bees, Bull. Unv. New Mexico, 1(1):41-73 (same paper as 1898b). 1898d. Synopsis of the North American bees of the genus Nomia, Ento- mologist, 31(417) :31-33. 1899a. The bees of Kansas, Ent. News, 10(1):3-4. 1899b. Catalogo de las abejas de México, Ofic. Tip. de la Secretaria de Fomento, México, pp. 1-20. 1900. Observations on bees collected at Las Vagas, New Mexico, and in the adjacent Mountains, Ann. Mag. Nat. Hist., (7) 5:401-416. 1902. Flowers and insects in New Mexico, Amer. Naturalist, 36(430): 809:817. 1905. Notes on some bees in the British Museum, Trans. Amer, Ent. Soc., 31:309-364. 1906a. The bees of New Mexico, Trans. Amer. Ent. Soc., 32:289-314. 1906b. Some bees from Washington State, Canadian Ent., 38:277-282. 1907. The bees of Boulder County, Colorado, Univ. Colorado Studies, 4(4) :239-259, 1908. Descriptions and records of bees—XX, Ann. Mag. Nat. Hist., (8) 2:323-334. 1910a. Descriptions and records of bees—XXXII, Ann. Mag. Nat. Hist., (8) 6:272-284. 1910b. The North American bees of the genus Nomia, Proc. U. S. National Mus., 38( 1745) :289-298. BANDED SUBGENERA OF NOMIA IN AMERICA 341 1912. Descriptions and records of bees—XLVII, Ann. Mag. Nat. Hist., (8) 10:484-494. 1914. Bees visiting Helianthus, Canadian Ent., 46( 12) :409-415. 1917. In W. P. Cockerell, Collecting bees in southern Texas, J. New York Ent. Soc., 25(3):187-193. 1918. Descriptions and records of bees—LXXIX, Ann. Mag. Nat. Hist., (9) 1:158-167. 1923. Expedition of the California Academy of Sciences to the Gulf of Califomia in 1921. The bees (1), Proc. California Acad. Sci., (4) 12(7):73-103. 1924. Expedition of the California Academy of Sciences to the Gulf of California in 1921. The bees (II), Proc. California Acad. Sci., (4) 12(27):529-560. 1925a. Bees in the collection of California Academy of Sciences, Proc. California Acad. Sci., (4) 14(11):185-215. 1925b. Descriptions and records of bees—CV, Ann. Mag. Nat. Hist., (9) 16:228-233. 1925c. Descriptions and records of bees—CVII, Ann. Mag. Nat. Hist., (9) 16:621-629. 1928. Supplementary notes on Colorado bees, with a list of all the genera, Univ. Colorado Studies, 16(2):99-126. 1930. In Cockerell and Blair, Rocky Mountain bees. I, Amer. Mus. Novitates, 433:1-7, 10-17. 1933. New or little-known western bees, Pan-Pac. Ent., 9(4):153-159 (see 1934a). 1934a. Pan-Pac. Ent., 10(1):21 (correction of 1933). 1934b. Records of western bees, Amer. Mus. Novitates, 697:1-15. COocKERELL, T. D. A. and W. W. Rossins 1910. An introduction to the study of Rocky Mountain bees, Univ. Colorado Studies, 7(3):179-195. CracIn, F. W. 1886. Hymenoptera collected in Barber County [Kansas], in late July and early August, 1886, Bull. Washburn College Lab. Nat. Hist., 1(17): 211. CRAWFORD, J. C. 1911. Descriptions of new Hymenoptera. I, Proc. U. S. National Mus., 39( 1804) :617-623. Cresson, E. T. 1865. On the Hymenoptera of Cuba, Proc. Ent. Soc. Philadelphia, 4:1-200. 1868. Catalogue of a small collection of Hymenoptera made in New Mexico during the summer of 1867, Trans, Amer. Ent. Soc., 1: 375-388. 1872. Hymenoptera Texana, Trans. Amer. Ent. Soc., 4:153-292. 1875. Report upon the collections of Hymenoptera made in portions of Nevada, Utah, Colorado, New Mexico, and Arizona during the years 1872, 1873, and 1874, Rept. Geog. Geol. Expl. Survey west of 100th Meridian, 5:703-728, plates 33-34. 14—1367 342 THE UNIVERSITY SCIENCE BULLETIN 1887. Synopsis of the Hymenoptera of American north of Mexico, Trans. Amer. Ent. Soc., supplementary vol. 1887, p. i-vi, 1-350. 1916. The Cresson types of Hymenoptera, Memoirs Amer. Ent. Soc., 1:1-141. Cross, EARLE A. 1958a. A revision of the bees of the subgenus Epinomia in the New World (Hymenoptera: Halictidae), Univ. Kansas Sci. Bull., 38(2): 1261-1301. 1958b. Notes on artificially induced nesting of the alkali bee, Nomia melanderi Ckll., Proc. Indiana Acad. Sci., 1957, 67:135-136 (ab- stract ). Cross, EarLE A. and Greorcr E. BOHART 1960. The biology of Nomia (Epinomia) triangulifera with comparative notes on other species of Nomia, Univ. Kansas Sci. Bull., 41: 761-792. Daa Torre, C. G. de 1896. Catalogus hymenopterorum hucusque descriptorum systematicus, vol. 10 Apidae (Anthophila), Leipzig, pp. i-viii, 1-643. Doutt, Kerr M. and J. D. Purvi 1960. Lucerne for seed, South Australia Dept. Agr., Bull., 463:1-23. Evans, Howarp E. and E. Gorton LiInsLEy 1960. Notes on a sleeping aggregation of solitary bees and wasps, Bull. S. California Academy Sci., 59(1):30-37, plate 12. Fattic, P. W. 1943. The Mutillidae or velvet ants of Georgia, Emory Univ. Mus. Bull., 1:1-24. Frick, KENNETH E. 1957. Biology and control of tiger beetles in alkali bee nesting sites, J. Econ. Ent., 50(4):503-504. 1962. Ecological studies on the alkali bee, Nomia melanderi, and _ its bombyliid parasite Hetrostylum robustum, in Washington, Ann. Ent. Soc. Amer., 55(1):5-15. Frick, KENNETH E., Howarp Porrer and HENry WEAVER 1960. Development and maintenance of alkali bee nesting sites, Wash. Agr. Expt. Sta. Circular, 366:1-10. FRIESE, HEINRICH 1897. Monographie der Bienengattung Nomia (Latr.) (Palaearctische Formen), In Fest-Schrift zur Feier des fiinfzigjahrigen Bestehens des Vereins fiir Schlesische Insektenkunde in Breslau, 1847-1897, pp. 45-84. Fronx, W. D. and L. I. PAInrer 1960. Some characteristics of alkali bee nesting sites, J. Econ. Ent., 53(3) :424-425, Fox, Wo. J. ' 1893. Synopsis of the N. American species of Nomia, Ent. News, 4(4): 134-135. BANDED SUBGENERA OF NOMIA IN AMERICA 343 HissBarp, CLAUDE W. 1960. An interpretation of Pliocene and Pleistocene climates in North America, 62d Ann. Rept. Michigan Acad. Sci., Arts and Letters, pp. 5-30. Hlicu, M. M. 1921. An unusual type of injury to sweet potatoes in Texas by a burrow- ing bee, J. Econ. Ent., 14(3):306-307. HirASHIMA, YOSHIHIRO 1961. Monographic study of the subfamily Nomiinae in Japan, Acta Hymenopterologica, 1(3):241-303. Howe Lu, J. FRANKLIN 1960. A technique employing bio-plastic in the excavation of burrows of ground-nesting bees, Ann. Ent. Soc. Amer., 53(5):679-682. Hoyt, Avery S. 1950. Report of the Chief of the Bureau of Entomology and Plant Quarantine 1950, U. S. Dept. Agr., Washington, D. C., pp. 1-72. 1951. Report of the Chief of the Bureau of Entomology and Plant Quaran- tine 1951, U. S. Dept. Agr., Washington, D. C., pp. 1-78. 1952. Report of the Chief of the Bureau of Entomology and Plant Quar- antine 1952, U. S. Dept. Agr., Washington, D. C., pp. 1-84. James, Maurice T. 1955. A new sarcophagid parasite of Nomia bees, Proc. Ent. Soc. Wash- ington, 57(6):283-285. KNowLEs, R. P. 1943. The role of insects, weather conditions, and plant character in seed setting in alfalfa, Sci. Agr., 24(1):29-50. KNOwLTON, GEORGE F. 1952. Growing alfalfa seed without destroying pollinators, Utah State Agr. College Extension Circular, 155:1-2. KROMBEIN, Kart V. 1953. The wasps and bees of the Bimini Island group, Bahamas, British: West Indies (Hymenoptera: Aculeata), Amer. Mus. Novitates, 1633:1-29. 1958. Hymenoptera of America north of Mexico, synoptic catalog, Ist supplement, Agr. Monog. 2, U. S. Dept. Agr., Washington, D. C., pp. 1-350. LAFFOON, JEAN L. 1960. Common names of insects, Bull. Ent. Soc. Amer., 6(4):175-210. LINsLEy, E. GorTON 1946. Insect pollinators of alfalfa in California, J. Econ. Ent., 39(1): 18-29. 1958. The ecology of solitary bees, Hilgardia, 27(19) :543-599. 1960. Ethology of some bee- and wasp-killing robber flies of southeastern Arizona and western New Mexico, Univ. California Publications in Ent., 16(7):357-392, plates 48-55. 1962. The colletid Ptiloglossa arizonensis Timberlake, a matinal pollina- tor of Solanum, Pan-Pac. Ent., 38(2):75-82. 344 Tue UNIVERSITY SCIENCE BULLETIN Linsury, E. G. and J. W. MacSwain 1947a. 1947b. Linstey, E 1950. The effect of DDT and certain other insecticides on alfalfa pollina- tors, J. Econ. Ent., 40(3):358-363. Factors influencing the effectiveness of insect pollinators of alfalfa in California, J. Econ. Ent., 40(3):349-357. . G., J. W. MacSwain and Ray F. SmirH Comparative susceptibility of wild bees and honey bees to DDT, J. Econ. Ent., 43(1):59-62. MENKE, HERMAN F. 1951. 1952a. 1952b. 1953. 1954a. 1954b. MICHENER, 1936. 1937. 1944. 1951. 1954. IO Gi MITCHELL, 1960. Toxicity of some insecticides to Nomia melanderi and Apis mellifera, J. Econ. Ent., 44(4) :624-625. Alkali bee helps set seed records, What’s New in Crops and Soils, 4(8):16-17. A six million dollar native bee in Washington state, Amer. Bee J., 92(8) :334-335. Alkali bee, Co-op Grain Quarterly [Co-op Grain Publishing Co., St. Paul], 11:15-17. Repellency of toxaphene dust and parathion spray to Nomia melanderi in blossoming alfalfa, J. Econ. Ent., 47(3):539-540. Insect pollination in relation to alfalfa seed production in Wash- ington, State College of Washington, Washington Agr. Expt. Sta. Bull., $i55:1-24. CuHarues D. ; Some western anthophorid and Nomiinae bees, Amer. Mus. Novitates, 876:1-4. Recor 1s and descriptions of North American bees, Ann. Mag. Nat. Hist. (10) 19:313-329. Com parative external morphology, phylogeny, and classification of the bees (Hymenoptera), Bull. Amer, Mus. Nat. Hist., 82(6): 151-326. In Muesebeck et al., Hymenoptera of America north of Mexico, synoptic catalog, Agr. Monog. 2, U. S. Dept. Agr., Washington, D. C., pp. 1043-1255. Comparative morphological and systematic studies of bee larvae with a key to the families of hymenopterous larvae, Univ. Kansas Sci. Bull., 35(2):987-1102. Observations on the pupae of bees, Pan-Pac. Ent., 30(1):63-70. Comments on some groups of Nomiinae, Acta Hymenopterologica, 1(3):239-240. THEODORE B. Bees of the eastern United States, vol. 1, North Carolina Agr. Expt. Sta. Tech, Bull., 141:1-538. Moose, GEORGE 1960. This is for the bees . . . also for larger alfalfa seed crops, Oregon Dept. Agr., Agr. Bull., 206:7. Movrg, J. S. and C. D,. MicHENER 1955. A contribution towards the classification of Neotropical Eucerini, Dusenia, 6(6) :239-331. BANDED SUBGENERA OF NOMIA IN AMERICA 345 MuESEBECK, C. F. W., et al. 1951. Hymenoptera of American north of Mexico, synoptic catalog, Agr. Monog. 2, U. S. Dept. Agr., Washington, D. C., pp. 1-1420, map. Nye, W. P. and G. E. Bouart 1959. Photographing insects close up, J. Biol. Photo. Assoc., November, pp. 139-145. Parks, H. B. 1928. Nesting habits of the bees, Nomis [sic] nortonii Cresson, in Texas, Bull. Brooklyn Ent. Soc., 23:263. Patton, W. H. 1879. List. of a collection of aculeate Hymenoptera made by Mr. S. W. Williston in northwestern Kansas, Bull. U. S. Geol. Geog. Survey of the Territories, 5(3):349-370. Rav, PHin 1943. Notes on the nesting habits of certain social and solitary bees of Mexico, Ann. Ent. Soc. Amer., 36(4):641-646. Ross, T. R. 1957. Bees. Wyoming alfalfa seed growers could not get along without these insect pollinators, Wyoming Roundup [Wyoming Agr. Expt. Sta.], 4(4):3-5. ROBERTSON, CHARLES 1892. Flowers and insects—Labiatae, Trans. Acad. Sci. St. Louis, 6(4): 101-181. 1904. Synopsis of Anthophila, Canadian Ent., 36(2):37-43. 1925. Heterotropic bees, Ecology, 6(4):412-436. 1928. Flowers and insects, Sci. Press Printing Co., Lancaster, Pennsy]l- vania, pp. 1-221. SAKAGAMI, SHOrcHI F. and CHarites D. MICHENER 1962. The nest architecture of the sweat bees, Univ. Kansas Press, Law- rence, pp. 1-135. SANDHOUSE, GRACE A. 1943. The type species of the genera and subgenera of bees, Proc. U. S. Nat. Mus., 92(3156) :519-619. SELANDER, RicHarp B. and Grorcr E. Bouart 1954. The biology of Zonitis atripennis flavida LeConte (Coleoptera: Meloidae), Wasmann J. Biol. [Univ. San Francisco], 12(2): 227-248. Snow, F. H. 1880. Preliminary list of the Hymenoptera of Kansas, Trans. Kansas Acad. Sci., 7:93-98. 1907. List of species of Hymenoptera collected in Arizona by the Univer- sity of Kansas Entomological Expeditions of 1902, 1903, 1904, 1905, and 1906, Trans. Kansas Acad. Sci., 20(2):127-139. SNYDER, Karz D. 1957. Check list of insects of Great Smoky Mountains National Park, pp. 1-78 (mimeographed ), 346 THe UNIvERSITY SCIENCE BULLETIN STEPHEN, W. P. 1959a. Maintaining alkali bees for alfalfa seed production, Oregon State College, Agr. Expt. Sta. Bull., 568:1-23. 1959b. The alkali bee and alfalfa seed, Proc. 19th Ann. Meeting Oregon Seed Growers League, pp. 69-70. 1960a. Studies in the alkali bee (Nomia melanderi Ckll.), II. Preliminary investigations on the effect of soluble salts on alkali bee nesting sites, Oregon State College, Agr. Expt. Sta. Tech. Bull., 52:15-26, 1960b. Studies in the alkali bee (Nomia melanderi Ckll.), III. Manage- ment and renovation of native soils for alkali bee inhabitation, Oregon State College, Agr. Expt. Sta. Tech. Bull., 52:27-39. 1960c. Artificial bee beds for the propagation of the alkali bee, Nomia melanderi, J. Econ. Ent., 53(6):1025-1030. 1960d. Soil requirements for alkali bee nesting sites, Rept. 17th Alfalfa Improvement Conf., pp. 30-31. STEPHEN, W. P. and D. D. Evans 1960. Studies in the alkali bee (Nomia melanderi Ckll.), I. Soil physical requirements for bee nesting, Oregon State College, Agr. Expt. Sta. Tech. Bull., 52:1-14. STEPHEN, W. P. and P. F. TorcHi1o 1961. Biological notes on the leaf-cutter bee, Megachile (Eutricharaea) rotundata (Fabricius), Pan-Pac. Ent., 37(2):85-93. STEVENS, O. A. 1951. Native bees. Furrow bees—Halictus, North Dakota Agr. Expt. Sta. Bimonthly Bull., 14(2):59-64. 1954. Wild bees and alfalfa pollination, North Dakota Agr. Expt. Sta. Bimonthly Bull., 17(1):32-34. SwENK, Myron H. 1907. The bees of Nebraska.—III, Ent. News, 18(7):293-300. Topp, FRANK E. 1957. Insect pollination of legumes, in Wheeler and Hill, Grassland Seeds, D. van Nostrand Co., Inc., Princeton, New Jersey, pp. 62-76. Topp, FRANK E. and S. E. McGrecor 1961. Insecticides and honey bees, in Seeds, Yearbook of Agr., U. S. Dept. Agr., Washington, D. C., pp. 247-250. TONZ, CLARENCE 1954. A hard-working bee in the west, Gleanings in Bee Culture, 82(7): 418-419. TyspaL, H. M. 1940. Is tripping necessary for seed setting in alfalfa?, J. Amer. Soc. Agronomy, 32(7):570-585. 1946. Influence of tripping, soil moisture, plant spacing, and lodging on alfalfa seed production, J. Amer. Soc. Agronomy, 38(6):515-535. VACHAL, J. 1897. Quelques espéces nouvelles, douteuses ou peu connues du genre Nomia Latr. (Hym.), Miscellanea Entomologia, 5(6):72-75, 5(7): 87-88, 5(8):89-93. BANDED SUBGENERA OF NOMIA IN AMERICA 347 VANSELL, GEORGE H. and Frank E. Topp 1946. Alfalfa tripping by insects, J. Amer. Soc. Agronomy, 38( 6) :470-488. 1947. Honeybees and other bees pollinate the alfalfa seed crop in Utah, Gleanings in Bee Culture, 75(3):136-139, 183. VANSELL, GEorcE H. and W. H. Griccs 1952. Honey bees as agents of pollination, in Insects, Yearbook of Agr., U. S. Dept. Agr., Washington, D. C., pp. 88-107. ViereEck, H. L. 1906. Notes and descriptions of Hymenoptera from the western United States, Trans. Amer, Ent. Soc., 32(2):173-247. WEstTWoop, J. O. 1875. Descriptions of some new species of short-tongued bees belonging to the genus Nomia of Latreille, Trans. Ent. Soc. London, 1875(3): 207-222, plates 4-5. WueEELer, W. A. and D. D. Hitt, editors 1957. Grassland Seeds, D. van Nostrand Co., Inc., Princeton, New Jersey, pp. i-xx, 1-734. WILLE, ALVARO 1956. Comparative studies of the thoracic musculature of bees, Univ. Kansas Sci. Bull., 38( 1) :439-499. 1958. A comparative study of the dorsal vessels of bees, Ann, Ent. Soc. Amer., 51(6) :538-546. 348 THE UNIVERSITY SCIENCE BULLETIN MAP 1 The distribution of Nomia (Curvinomia) maneei, N. (C.) universitatis, N. (C.) t. tetrazonata and N. (C.) t. uvaldensis. BANDED SUBGENERA OF NOMIA IN AMERICA 349 14a—1367 350 THE UNIVERSITY SCIENCE BULLETIN © N foxii BN fedorensis MAP 2 The distribution of Nomia (Curvinomia) foxii and N. (C.) fedorensis. BANDED SUBGENERA OF NOMIA IN AMERICA 3851 @ N. angustitibialis MAP 3 The distribution of Nomia (Curvinomia) angustitibialis. 352 Tue UNIversiry SCIENCE BULLETIN MAP 4 The distribution of the species and subspecies of Nomia (Acunomia). The probable range of N. (A.) melanderi howardi is shaded to correspond with the western range of Larrea in the United States and with the western desert of Mexico. Hemet and La Quinta are indicated by a and b respectively in the range of N. (A.) melanderi. The shading of N. (A.) nortoni cressoni is used only to emphasize its range. 393 BANDED SUBGENERA OF NOMIA IN AMERICA INOSNIGOW NY 1QUVROH AN 1W20NVI3IA Ay INO6G3¥9 WN, IWOLUON WIRY 354 Tue UNIVERSITY SCIENCE BULLETIN FIGURES 1-11 Fics. 1-11. Nomia (Curvinomia) and N. (Acunomia) species, Right hind tibia, male, side view. N.. (C.) maneei. N. (C.) fedorensis. N. (C.) universitatis. N. (C.) angustitibialis. N. (C.) foxii. N. (C.) tetrazonata uvaldensis N. (A.) n. nortoni. N. (A.) n. cressoni. . N. (A.) m. melanderi. 10. N. (A.) m. howardi. 11. N. (A.) robinsoni. The area below the dotted line is lighter in color and translucent. Sere 2) Nien) SEM hee IS) lee BANDED SUBGENERA OF NoMIA IN AMERICA 355 FIGURES 1-11 356 THe UNIVERSITY SCIENCE BULLETIN FIGURES 12-28 Fics. 12-17. Nomia (Curvinoma) species. Metasomal sternum 5, male, ventral aspect, right side only. 12. 13. 14. 15. 16. ule N. (C.) maneei. N. (C.) fedorensis. N. (C.) universitatis. N. (C.) angustitibialis, with enlargements of hairs. N. (C.) foxii. N. (C.) t. tetrazonata. Fics. 18-23. Nomia (Curvinomia) species. Left penis valve, ventral view; above, tip of penis valve, side view. 18. 19. bo S WN w by one N. (C.) maneei. N. (C.) fedorensis, N. (C.) universitatis. N. (C.) angustitibialis. N. (C.) foxii. N. (C.) t. tetrazonata, BANDED SUBGENERA OF NOMIA IN AMERICA 357 IN ANY , RY \\, 4 if WY RANA NY mys iN WY TAN. NAT MW VHS + Me 15 MWe reg pms v FIGURES 12-23 M7 W144 As EAS, SRN Se. ag ae ec Ue, 14 il ean 3 @ IN Ri Say es N fo) OI SSS 23 358 ‘} HE UNIVERSITY SCIENCE BULLETIN FIGURES 24-32 Fics. 24-26. Nomia (Acunomia) species. Metasomal sternum 5, male, ventral aspect. 24. N. (A.) n. nortoni, the letters refer to structures discussed in the text: a. median, basal emargination, b. fused row of setae, c. lateroapical hairy pad, d. median, apical process, e. median, longitudinal ridge. 25. N. (A.) m. melanderi, with enlargements of hairs. 26. N. (A.) robinsoni. Fics. 27-29. Nomia (Acunomia) m. melanderi. Metasomal sternum 6-8, male, ventral aspect, right side only. Fig. 27. Stemum 6. Fig, 28. Stemum 7. Fig. 29. Sternum 8. Fics, 30-32. Nomia (Acunomia) species. Genitalia, left side only. 30. N. (A.) n. nortoni, ventral view, the letters refer to structures discussed in the text: a. gonobase, b. ventroapical, median part of gonobase, c. gono- coxite, d. dorsal recurved part of gonocoxite, e. gonostylus, f. hooklike process of gonostylus, g. basal plate of gonostylus, h. penis valve, i. anteroventral process of penis valve, j. ventral opening of penis valve, k. lateral part of penis valve, 1. inner subapical margin of penis valve, m. tip of penis valve. 31. N. (A.) m. melanderi, ventral (left) and dorsal view (right). 32. N. (A.) robinsoni, ventral view. BANDED SUBGENERA OF NOMIA IN AMERICA 359 FIGURES 24-32 LP 4 7) //' 1. ZENG SS u — ————< THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Wis 2G) JuNE 7, 1965 [No. 4 A Taxonomic Study of the Subgenus Ladona (Odonata: Libellulidae ) * BY Bitiy L. BENNEFIELD INTRODUCTION This paper is the result of a taxonomic study of the dragonfly subgenus Ladona in the Nearctic region. Some material concern- ing the biology of the group is also included, as well as a study of variation within one species, Libellula (Ladona) julia. Some 1,200 adults and 15 nymphs of the subgenus were examined. Prior to this work, the taxonomy of this group has been based largely on coloration, and descriptions were based on relatively few specimens from one or two localities. For this reason, empha- sis has here been placed on the recognition of taxonomic characters other than coloration, and an attempt has been made to re-describe the species more exactly on the basis of these characters as well as coloration. TECHNIQUES AND TERMINOLOGY The characters and terminology used in the keys are, for the most part, self-explanatory or are clarified by illustrations. However, to facilitate accurate and rapid interpretation of descriptions, certain techniques and terms are explained. All measurements listed are in millimeters. Gross body measure- ments (e. g., wing length) were made with a standard millimeter rule. Measurements requiring finer divisions (e. g., stigma) were made with an ocular micrometer. The following measurements were made: * Contribution No. 1187 from the Department of Entomology, The University of Kansas, Lawrence, Kansas; submitted in partial fulfillment of the requirements for the degree of Master of Arts. Published posthumously. (361) 15—1367 362 THE UNIVERSITY SCIENCE BULLETIN 1. Total length—maximum mid-dorsal, from front of head to distal ends of anal appendages, to nearest millimeter. 2. Abdomen—maximum mid-dorsal length from anterior transverse carina of segment I to distal ends of anal appendages. 3. Fore wing—maximum length from tip to base. 4. Hind wing—maximum length from tip to base. 5. Stigma—along costa of fore wing. 6. Length nodus to stigma—maximum along costa from nodus to proximal edge of stigma of fore wing. 7. Metathoracic femur—maximum length along lateral meson. 8. Front, middle, and hind tibiae—maximum length along lateral meson. 9. Head width—maximum, dorsal view at eye level. 10. Superior anal appendage—maximum mid-lateral length. 11. Supra-anal plate—maximum length in lateral view. The size of all structures drawn is indicated in millimeters and is given as actual size. Color, which was used almost entirely in the descriptions of earlier workers, has been found to be variable and in some cases of reduced value in species recognition. There is color variation both with age and geographical distribution. Few, if any, individuals will be found to agree in every detail with the original color descriptions. An attempt has been made to treat coloration in the species de- scriptions in such a way as to include as much of the range of variation as possible. In order to facilitate a more accurate description of coloration of the lateral surface of the synthorax, the individual sclerites are considered. All color determinations were made using low (15 x ) magnification with an unfiltered light source directed from above The coloration of the wings is relatively stable and reliable within species. In order to discuss this coloration, the term inter- space has been used (e. g., costal-subcostal interspace). This refers to the space between the two veins named. While the female genitalia were found to be similar in all three species and thus of little use in species determination, those of the male, especially the penis and posterior hamules, were found to be specifically distinct. Observation of the penis usually necessitates its removal from the specimen. For observations of the other char- acters used in this paper, clearing is not necessary. The morpho- logical terminology used for the description of the penis is that of Chao (1953) and differs somewhat from the terms used by Kennedy (1922). Other morphological terminology has also been adopted from Chao (1953). A TAXONOMIC STUDY OF THE SUBGENUS LADONA 363 DESCRIPTION AND TAXONOMIC HISTORY OF THE SUBGENUS LADONA NEEDHAM 1897 Libellula, Williamson, 1900, Rpt. Indiana Dept. Geol., vol. 24, p. 328; Calvert, 1906, Biologia Centrali-Americana, Neuroptera, p. 206; Williamson, 1922, Ent. News, vol. 33, p. 18; Byers, 1927, Ent. News, vol. 38, pp. 113-114; Calvert, 1927, Univ. Indiana Stud. Nat. Hist., vol. 12, p. 2; Byers, 1930, Univ. Florida Pub. Biol. Sci., vol. 1, pp. 106, 111; Byers, 1936, Ent. News, vol. 47, p. 62. Ladona, Needham, 1897, Canadian Ent., vol. 29, pp. 144-149; Needham, 1901, Bull. New York St. Mus., vol. 47, pp. 507-508, 528; Muttkowski, 1908, Bull. Wisconsin Nat. Hist. Soc., vol. 6, pp. 105, 115; Needham and Hey- wood, 1929, Handbook of North American Dragonflies, pp. 202, 204, 217; Needham and Fisher, 1936, Tran. American Ent. Soc., vol. 62, pp. 113, 115; Brimley, 1938, Insects of North Carolina, p. 39; Wright and Peterson, 1944, Ohio Jour. Sci., vol. 44, pp. 158, 165; Needham and Westfall, 1955, Dragon- flies of North America, pp. 58, 426, 428, 431-432, 473; Smith and Pritchard, 1956, Aquatic Insects of California, p. 124. Libellula (Ladona), Borror, 1945, Ann. Ent. Soc. America, vol. 38, pp. 171, 174, 175, 192; Ris, 1910, Coll. de Selys, vol. 10, p. 245; Kennedy, 1922, Ent. News, vol. 33, pp. 105, 109, 110. Type Species: Libellula exusta Say, 1839. This subgenus is composed of dragonflies of medium size, which may be readily separated from the other members of the genus Libellula by vein M,, which arises before the middle of the stigma. The wings are clear with basal brown spots. There is usually one crossvein present in the super triangle, and usually in the anal area behind vein A, there are three well defined rows of cells. The abdomen gradually tapers posteriorly and is relatively uni- form in coloration in exusta and deplanata, while julia usually bears white coloration on the first four segments. The posterior lobe of the middle segment of the penis is con- siderably more reduced, and the cornula of the penis is relatively shorter and more inconspicuous than in other members of the genus Libellula. The members of this group are associated with ponds. They are rapid flyers and fly quite close to the surface of the water. They exhibit a preference for bare, sunny areas when at rest. The female lays her eggs while hovering over the surface of the water. The known nymphs may be readily separated from the other nymphs of Libellula on the basis of the crenulate median lobe of the labium. All other nymphs of the genus Libellula have the median lobe of the labium smooth and evenly contoured. The members of this subgenus range from the southeastern Canadian provinces through some of the central and southern states to Florida, westward to Washington (Fig. 1). 364 THE UNIVERSITY SCIENCE BULLETIN Three species form this subgenus. The status of these three species was confused by many early authors. Hagen (1861) con- sidered deplanata as a synonym of exusta and later (1875) con- sidered deplanata as a southern form of exusta and julia as a syno- nym of exusta. Calvert (1893) placed both julia and deplanata as synonyms of exusta. The three species were then considered under the name exusta until 1897. At this time, Needham (1897) recognized all three species and erected the genus Ladona to con- tain them, with exusta as the type species. Due to the uniqueness of the type species of Libellula, Needham felt that all American species of the genus Libellula should be placed in other genera, hence the proposal of the genus Ladona. It is now generally agreed that this step was not necessary. The genus Ladona was not well received, and many workers such as Ris (1910), Garman (1927), Byers (1930), and others continued to place its species in Libellula. In addition, the taxonomic rank of the three forms continued in dispute. For example, Muttkowski (1910) considered julia as a synonym of exusta and deplanata as a subspecies of julia. Byers (1930) considered julia as a synonym of exusta and recognized only two species, exusta and deplanata. Several authors have considered the three species to form a sub- genus of Libellula. For example, Kennedy (1922) developed a phylogeny for the genus Libellula on the basis of the morphology of the penis and placed Ladona as a subgenus. The results of Kennedy’s study agree largely with those of Ris (1910). In view of these varied opinions, it seemed worthwhile to re-examine the evidence in order to form a sound judgment as to the status of Ladona. In order to accomplish this, two species of Libellula (semifasciata and pulchella) and one of Ladona (julia) were chosen. It was felt that a comparison of these three species would indicate the degree of relationship between the two groups. Li- bellula semifasciata Burmeister was chosen since it is considered to represent the most primitive species of Libellula (Kennedy, 1922). Libellula pulchella Drury was chosen as a typical representative of the genus Libellula. Ladona julia was chosen since it appears to be the most primitive of the three species of Ladona (Kennedy, 1922). Among the many morphological features studied, only the following differences were noted: 1. The epistomal suture is quite convex in semifasciata, slightly convex in pulchella, and virtually straight in julia. 2. In semifasciata the anterior margin of the median lobe of the labium is broadly pointed; in pulchella this structure is much less pointed, and in julia it is evenly contoured. A Taxonomic STuDY OF THE SUBGENUS LADONA 365 3. Libellula semifasciata exhibits a distinct notch or break in the middle lobe of the prothorax; this is less apparent in pulchella and absent in julia. 4. The posterior margin of the posterior lobe of the prothorax in julia was observed to have a fan-shaped expanse of hairs not found in either semi- fasciata or pulchella. 5. The dorsal lobe of the eucervicale is somewhat more rounded in semi- fasciata than in pulchella or julia. 6. The mesopleural suture is less sigmoid in julia than in either semifasciata or pulchella. 7. The metapleural suture is straight in semifasciata, sigmoid in pulchella and julia. On the basis of these comparisons, the placement of Ladona at the subgeneric level seems most appropriate. After a study of wing venation, Borror (1945) reached the same conclusion and also treated Ladona as a subgenus. Key to Males of the Subgenus Ladona 1. Dorsum of anepisternum II without lateral antehumeral stripe; abdomi- nal segments IV-IX laterally black with irregular brown ventral spots, or wholly black; subcostal-radial and medial interspace of hindwing dark brown to level of second or third antenodal crossvein; radial and medial-postcubital interspace of both wings marked faintly with brown (Fig. 8); posterior ventral margin of segment X concave........ julia Dorsum of anepisternum II with two lateral antehumeral stripes; abdomi- nal segments IV-IX wholly black or with dorsal black triangles; sub- costal-radial and medial interspace of hindwing dark to a level beyond third antenodal crossvein; radial and medial-postcubital interspace clear or faintly marked with brown; posterior ventral margin of seg- MEP PC Straight ror CONCAVE acts 2c epee eC ae ee ee . Dorsum of anepisternum II brown with two yellow antehumeral stripes; abdominal segments IV-IX with dorsal black triangles; subcostal- radial and medial interspace of hindwing dark brown to level of fourth or fifth antenodal crossvein; radial and medial-postcubital interspaces of both wings clear (Fig. 9); posterior ventral margin of segment X SLAIN ers en he eae eternal Sats eure mnanne ee aria deplanata Dorsum of anepisternum IIT medium brown with two yellow-gray ante- humeral stripes; abdominal segments wholly black in mature speci- mens; subcostal-radial and medial interspace of hindwing dark brown to level of fifth or sixth antenodal crossvein (Fig. 10); radial and medial-postcubital interspaces of both wings marked faintly with brown; posterior ventral margin of segment X concave........ exusta to i) key to Females of Subgenus Ladona 1. Dorsum of anepisternum II uniform cream to brown; antehumeral stripes EN OFS SH TLS ks seseeigaw oye aN eas Aen oan er Se a aaa Beat eb ca qulia Dorsum of anepisternum II light brown to brown; antehumeral stripes DECSEMIUER, ss55 Sih naiet ew ghcegeinind Nexans Sin ebapne ns onal opan ete ler aaa t tegmnrmenr ey . Dorsum of anepisternum II brown with two cream antehumeral stripes, darkening with age; radial and medial-postcubital interspace of both wings; clearnGHigs9))asbicien 8 ckeycmmame ea. nu ah anes, aia Nisa ak deplanata Dorsum of anepisternum II light brown with two gray antehumeral stripes; radial and medial-postcubital interspace of both wings marked with Lorn Aew Miirez es KON st iN id Roney Ace Nt ce Soa Auer nei taiaese me gUieie ie) ai exusta bo i) 366 THE UNIVERSITY SCIENCE BULLETIN Key to Known Nymphs of Subgenus Ladona 1, Mental ‘setae three somes ee rani Pee ee Ron dane aes julia Mentalisetae absentinic iin eee Se On eee ee deplanata Libellula (Ladona) julia Uhler Libellula julia Uhler, 1857, Proc. Acad. Nat. Sci. Philadelphia, p. 88; Hagen, 1861, Synopsis of Neuroptera of North America, p. 153; Kirby, 1890, Syn- optic Catalogue of Neuroptera-Odonata, p. 26; Calvert, 1893, Trans. Ameri- can Ent. Soc., vol. 20, p. 258; Calvert, 1905, Occ. Pap. Boston Soc. Nat. Hist., vol. 7, p. 34; Walker, 1906, Canadian Ent., vol. 38, p. 153; Walker, 1906, Rpt. Ent. Soc. Ontario, vol. 36, p. 70; Williamson, 1907, Ohio Nat., vol. 7, p. 150; Muttkowski, 1908, Bull. Wisconsin Nat. Hist. Soc., vol. 6, p. 166; Walker, 1915, Rpt. Ontario Dept. Marine Fish., vol. 47 (supple- ment), pp. 51, 59, 91; Howe, 1920, Mem. Thoreau Mus., vol. 2, pp. 71, 97, 101; Howe, 1921, Mem. Thoreau Mus., vol. 2 (supplement), idp Ils Howe, 1921, Proc. Boston Soc. Nat. Hist., vol. 36, 128; Walker, 1928, Univ. Toronto Stud. Biol. Sci., vol. 32, p. 48; Wellies 1929, Ent. News, vol. 40, p. 243; Walker, 1932, Univ. Toronto Stud., Pub. Ontario Fish Res. Lab., vol. 48, pp. 229, 244; Walker, 1933, Canadian Ent., vol. 65, p. 69; Walker, 1934, Rpt. Que Soc. Prot., vol. 26, p. 11; Walker, 1940, Canadian Ent., vol. 72, pp. 32-88; Walker, 1941, Trans. Roy. Canadian Inst., vol. 23, pp. 205, 206, 216, 2538; Walker, 1942, Proc. Nova Scotia Inst. Sci., vol. 20, p. 173. Ladona julia; Needham, 1901, Bull. New York St. Mus., vol. 47, pp. 520, 529, 530, 531; MacGillivray and Houghton, 1903, Ent. News, vol. 14, p. 263; Needham, 1903, Bull. New York St. Mus., vol. 68, pp. 246, 247; Needham and Anthony, 1903, Jour. New York Ent. Soc., vol. 11, p. 122; Osburn, 1905, Ent. News, vol. 16, p. 195; Byers, 1927, Ent. News, vol. 38, p. 114; Needham and Heywood, 1929, Handbook of Dragonflies of North America, pp. 217, 218; Byers, 1930, Univ. Florida Pub. Biol. Sci., vol. 1, pp. 111, 112; Borror, 1935, Ohio Jour. Sci., vol. 35, p. 454; Borror, 1937, Ohio Jour. Sci., vol. 37, p. 190; Borror, 1940, Ent. News, vol. 51, p. 78; Ahrens, 1941, Ent. News, vol. 52, p. 287; Montgomery, 1941, Proc. Indiana Acad. Sci., vol. 50, p. 250; Montgomery, 1943, Ent. News, vol. 54, p. 3; Gillespie, 1945, Ent. News, vol. 56, p. 62; Montgomery, 1945, Proc. Indiana Acad. Sci., vol. 54, pp. 218, 221, 223; Bromley, 1948, Proc. Ent. Soc. Washington, vol. 50, p. 237; Whitehouse, 1948, Trans. Roy. Canadian Inst., vol. 27, p. 37; Price, 1950, Ohio Jour. Sci., vol. 50, p. 37; Needham and Westfall, 1955, Dragonflies of North America, pp. 58, 473-475, 477. Libellula (Ladona) julia; Borror, 1945, Ann. Ent. Soc. America, vol. 38, pp. 171, 174, 175, 192; Ris, 1910, Coll. de Selys, vol. 11, p. 248; Kennedy, 1922, Ent. News, vol. 33, pp. 68-70, 111. Types. Male holotype, located in the Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts. Type with most of right forewing missing, the tips of both hind wings and both hind tarsi missing. Mate. Measurements: Total length, 39.29 + 1.558 mm. (holo- type 40.50 mm.); abdomen length, 24.50 + 1.024 mm. (holotype 26.00 mm.); forewing length, 31.93 + 1.888 mm. (holotype 30.50 m.); hind wing length, 30.82 + 1.142 mm. (holotype about 30.40 mm.); stigma length, 3.18 + .1557 mm.; length nodus to stigma, 11.66 + .655 mm. (holotype 11.60 mm.); metathoracic femur length, 6.23 + .282 mm. (holotype 6.6 mm.); prothoracic tibia length, 4.43 A TAXONOMIC STUDY OF THE SUBGENUS LADONA 367 + .244 mm. (holotype 4.80 mm.); mesothoracic tibia length, 4.78 + .265 mm. (holotype 5.30 mm.); metathoracic tibia length, 6.21 + 318 mm. (holotype 6.60 mm.); head width, 6.93 + .196 mm. (holotype 6.70 mm.); superior anal appendage length, 1.97 + .114 mm. (holotype 2.0 mm.); supra-anal plate length, 1.36 + .104 mm. (holotype 1.50 mm. ). Head with parietal region tan; compound eyes dark reddish brown; vertex black with punctate appearance; frons peripherally black, becoming tan ventrally with central portion cream, becoming uniformly black with age; epistomal suture appearing as gray or black straight groove; postclypeus light tan with anterior tentorial pits visible as light yellow patches, becoming dark with age; ante- clypeus tan to light brown, becoming darker with age; labrum orange with occasional darker median area; labium light yellow to orange with median lobe partially or wholly black, lateral lobes orange often with varying amounts of black. Prothorax with dorsal margin of anterior lobe gold, remainder black; middle lobe black with median ridge brown, bearing white hairs; posterior lobe black, bearing long white hairs laterally and posteriorly in fan-like fashion; dorsal lobe of eucervicale with anterior half gold, remainder black. Synthorax with acrotergite II black; mesostigmal plates black; area posterior to pit of tergal apophysis II black becoming light brown; collar carina light brown; dorsal carina black bordered by brown stripes; antealar carina black; antealar sinus white; anepis- ternum II dorsally white, laterally a uniform dark brown; kate- pisternum II with dorsal one-third dark brown as anepisternum II, remainder light brown; mesopleural suture deepened dorsally and ventrally to form deep grooves surrounded by blue-black spots; epimeron II light brown with occasional traces of darker brown dorsally; anepisternum III light brown; katepisternum III light brown; epimeron III light brown; metapleural suture deepened dorsally and ventrally to form deep grooves surrounded by black spots. Forewing with subcostal-radial and medial interspace dark brown to level of first antenodal crossvein, occasionally to second ante- nodal crossvein; radial and medial-postcubital interspace marked very faintly with light brown; postcubital-anal interspace dark brown, color extending approximately half way to anal crossing; stigma light tan, subtending two to three cells; antenodal cross- veins 12 to 16; postnodal crossveins 9 to 18. Hind wing with 368 THe UNIVERSITY SCIENCE BULLETIN subcostal-radial and medial interspace dark brown, the color ex- tending to level of second antenodal crossvein, occasionally to level of third antenodal crossvein; radial and medial-postcubital interspace marked with light brown; postcubital-anal interspace very dark brown, the color extending approximately half way to anal crossing; first anal interspace with dark brown triangle; stigma light tan, subtending two to three cells; antenodal crossveins 10 to 13; postnodal crossveins 9 to 14 (Fig. 8). Legs with femora black, occasionally becoming brown dorsally; tibiae black with yellow stripes on outer surface; tarsi black. Abdomen with segment I uniform light brown; segments I-IV white dorsally and laterally, rarely becoming darkened; segments V-VIII black dorsally and laterally, usually with irregular light brown spots on ventral region of lateral surface; segment IX black dorsally and laterally, gonopophyses light brown; segment X black dorsally and laterally, posterior ventral margin convex; supple- mentary transverse carinae black; posterior transverse carinae black with black spines; submarginal ventral carinae black; ventral ca- rinae light brown; acrotergites white to light brown; ventral surfaces of segments tan to light brown. Superior anal appendages brown darkening with age, pointed; inferior anal laminae widely separated. Penis with stem somewhat constricted, posterior projection sharply rounded; posterior margin of middle segment smooth; distal segment a broad pedicel; cornula sharply pointed posteriorly, broadly rounded anteriorly (Fig. 2). Posterior hamuli bifurcate with anterior process bearing a single laterally directed hook; lateral lobe sharply rounded (Fig. 3). FeMALe. Measurements: Total length, 37.45 + 1.616 mm.; ab- domen length, 23.86 + 1.316 mm.; forewing length, 32.0 = 1.663 mm.; hind wing length, 30.85 + 1.625 mm.; stigma length, 3.61 + .883 mm.; length nodus to stigma, 11.52 + .589 mm.; metathoracic femur length, 5.73 + .527 mm.; prothoracic tibia length, 4.15 + .866 mm.; mesothoracic tibia length, 4.56 + .986 mm.; metathoracic tibia length, 6.28 + .215 mm.; head width, 6.84 + .725 mm.; superior anal appendage length, 1.04 + .454 mm. Head with parietal region light tan; compound eyes as in male; vertex with dorsal half dark tan, ventral half black, becoming wholly dark with age; ocelli surrounded by light yellow border; frons uniform tan with occasional yellow spots, darkening with age; epistomal suture appearing as a straight colorless groove; A TAxONOmMIc STUDY OF THE SUBGENUS LADONA 369 postclypeus tan with ventral margin occasionally yellow, becoming greenish brown with age; anteclypeus as postclypeus, occasionally marked with irregular yellow lines; labrum yellow with darker median area; labium as in male. Prothorax with anterior lobe as in male; middle lobe as in male but with median ridge more pronounced; posterior lobe as in male but with longer hairs; dorsal lobe of eucervicale with more exten- sive gold coloration than in male. Synthorax with acrotergite, mesostigmal plates, area posterior to pit of tergal apophysis II, and collar carina as in male; dorsal carina black, with faint stripe dorsally; antealar carina tan; antealar sinus brown to tan; anepisternum II dorsally cream with occasional irregular brown areas, becoming brown with age; anepisternum II laterally dark brown as in male; remainder of synthorax as in male. Forewing with costal-subcostal interspace very light brown, the color extending to level of first antenodal crossvein; subcostal- radial and medial interspace dark brown to level of first or second antenodal crossvein; radial and medial-postcubital interspace very light brown as in costal-subcostal interspace, more extensive and slightly darker than in male; postcubital-anal interspace very dark brown extending to anal crossing; stigma brownish, darker than in male, subtending two to three cells; antenodal crossveins 13 to 16; postnodal crossveins 10 to 13. Hindwing with costal-subcostal interspace very faint brown to level of first antenodal crossvein; subcostal-radial and medial interspace dark brown to level of third antenodal crossvein; radial and medial-postcubital interspace as in forewing, more extensive than in male, extending to arculus; post- cubital-anal interspace dark brown to a point approximately half way beyond anal crossing; first anal interspace with brown triangle somewhat larger than in male; stigma as in forewing; antenodal crossveins 10 to 18; postnodal crossveins 9 to 13. Legs as in male. Abdomen with segment I dorsally black, becoming tan laterally and ventrally; segments II-III tan to brown dorsally, light tan laterally and ventrally; segments IV-IX with median dorsal black stripe, laterally tan to light brown with dorsal and ventral black margins, tan or light brown areas in these segments being gradually reduced in each segment, with segment IX almost wholly black; spines of posterior transverse carinae progressively more pronounced posteriorly; segment X with dorsal half black, ventral half light brown; supplementary transverse carina black; posterior transverse 370 THe UNIVERSITY SCIENCE BULLETIN carinae black; submarginal ventral carinae black; ventral carinae black; acrotergite light brown; antecostal sutures yellow or black. Anal appendage brown to black, elongate with broadly rounded apex, shorter than segment X; supra-anal plate tan to brown with broadly rounded apex, shorter than anal appendage. Nympu. Total length, 20-22 mm.; mentum with prominent me- dian projection; mental setae three; dorsal abdominal hooks over- lapping less than half of following segment; length of anal append- age 2.5 mm.; length of abdominal segments [X-X 2.0 mm. Bronomics. Needham (1897) reported this species to frequent small ponds and the borders of adjacent woods. Kormondy (1958) reported julia to be virtually ubiquitous in Michigan with a wide range of aquatic habitats. Needham stated that this species has leisurely flight with hovering and darting movements, whereas both Gillespie (1945) and Borror (1940) have indicated it to be a fast flier and difficult to catch. Gillespie reported having observed the species flying swiftly over water, close to the surface in an “un-Libel- lula” fashion. The males are found in low places around ponds, rest- ing on the ground with wings aslant and forward. The females ovi- posit while hovering over the water, otherwise remaining largely in the woods surrounding the ponds. Osburn (1905) reported this species in the late afternoon resting on stones in large numbers, some distance from a lake, with their wings outspread against the sides of the rocks, and when disturbed they would fly to other rocks and assume the same position. Osburn pointed out that due to their coloration, these dragonflies were inconspicuous except where the rocks had been blackened by recent forest fires. Stumps and tree roots seemed to have no attraction as resting sites. Need- ham (1897, 1901) indicated that the nymphs of julia are active, resting with the abdominal appendages widely spread apart, and transform in association with Tetragoneuria. Distripution. Type locality: “Puget Sound,” Washington. Con- necticut: Litchfield Co. (74, 19) (20 June-5 July); Tolland Co. (176,22) (14 June). Inviana: Wells Co. (128 ) (28 May); Elkhart Co. (346, 39) (20 May). Maine: Androscoggin Co. (1¢) (8 July); Kennebec Co. (28,12) (26 June); Lincoln Co. (178, 492) (11 June-30 July); Penobscot Co. (12) (29 June). MassacuusetTts: Barnstable Co. (14 ) (24 May); Hampden Co. (13, 22) (20-22 June); Middlesex Co. (108, 42) (29 May-14 July); Norfolk Co. (54, 19) (8 June-10 July); Worcester Co. (44, 22) (2 July). Micatcan: Alger Co. (16, 1¢) (July); A TAxoNomic STUDY OF THE SUBGENUS LADONA 371 Baraga Co. (1¢ ) (5 July); Charlevoix Co. (54,29) (15 June- 5 July ); Cheboygan Co. (226,72) (29 May-29 July); Chippewa Co. (26) (20-21 July); Clare Co. (128) (8 July); Clinton Co. (1¢@ ) (no date); Dickinson Co. (16,12) (17 July); Emmet Co. (26, 22) (21 May-7 July); Genesee Co. (18, 12) (25 June); Gladwin Co. (4 4 ) (16-22 June); Gogebic Co. (14,12) (3 July); Grand Traverse Co. (104,52) (27 June); Houghton Co. (34 ) (14 June-8 July); Ingham Co. (32) (10 May); Iosco Co. (24, 49) (21 July); Iron Co. (36,29 ) (17 June); Kent Co. (16,19 ) (30 May); Keweenaw Co. (76, 52) (11 June-5 July); Living- ston Co. (64, 132 ) (2-10 June); Luce Co. (68,792 ) (26 June); Mackinac Co. (124,92) (14 June-28 July); Marquette Co. (54, 1¢) (23 June-18 July); Mason Co. (14 ) (25 July); Menominee Co. (56, 12) (18 June-12 July); Midland Co. (14 ) (25 June); Missaukee Co. (346, 49) (16 May-14 July); Monroe Co. (14, 1¢) (26 June); Montmorency Co. (134,32) (28 June-7 July); Muskegon Co. (66) (26 June-l1 July); Oakland Co. (184, 52?) (22 May-26 June); Ogemaw Co. (14) (26 June); Oscoda Co. (36, 12) (28-25 June); Otsego Co. (86, 59) (24 June- 5 July); Presque Isle Co. (96, 39) (13 June-3 July); Roscom- mon Co. (48, 22) (28 June); Schoolcraft Co. (54, 59) (2 June-16 July); Shiawasee Co. (24, 12) (6 June); St. Joseph Co. (16,22) (15 May); Washtenaw Co. (224, 129) (11 May- 16 June); Wayne Co. (12) (no date). Minnesota: Anoka Co. (1¢) (13 June); Beltrami Co. (22 ) (16 June); Cass Co. (2¢ ) (8-15 June); Hennepin Co. (14, 392) (4 June-17 July); Hub- bard Co. (24 ) (5 June); Itasca Co. (114,792 ) (29 May-16 June); Ramsey Co. (96,22 ) (4 June-10 July); St. Louis Co. (34, 3¢ ) (21 May); Wabasha Co. (26 ) (22 June-18 July). New Hamp- sHiRE: Carroll Co. (44 ) (19 June-12 July); Cheshire Co. (14 ) (June); Grafton Co. (26,12) (no date). New Jersey: Bergen Co. (146) (no date); Morris Co. (34, 12) (28 May-22 June); Ocean Co. (24) (6 June); Sussex Co. (14) (31 May). New York: Duchess Co. (14 ) (22 May); Essex Co. (48) (9 July); Franklin Co. (36, 29 ) (16 June-25 July); Greene Co. (14,19 ) (28 May); Herkimer Co. (1¢,1¢2) (26 June); Hudson Co. (14, 1¢) (no date); Monroe Co. (146 ) (15 June); Nassau Co. (82, 29) (26 June); Onondaga Co. (26, 29) (7-15 June); Orange Co. (16,192) (9 July); St. Lawrence Co. (12 ) (19 June); Tioga Co. (1é) (4 July); Tompkins Co. (2?) (no date); Warren Co. (66, 32) (July); Washington Co. (14) (16 July). Onto: By) THe UNIVERSITY SCIENCE BULLETIN Williams Co. (4é ) (830 May). PENNsytvanra: Northampton Co. (26) (10 June). VeERMontT: Caledonia Co. (1é¢) (16 July). WasHINGTON: Spokane Co. (1é) (4 July); Thurston Co. (1@ ) (no date). Wrusconstn: Ashland Co. (1¢@ ) (3 June). In addition to the above material, the following Canadian ma- terial was examined. These localities are not included on the distribution map (Fig. 1). Bririsa Cotumpia: Vancouver Island (28, 292) (7-19 July); Nanino, Langford Lake (22 ) (20 July). New Brunswick: St. John’s (12) (18 June). Nova Scotia: Hants, Mt. Uniacke (2¢, 42) (no date); Shelburne, Allendale (14) (no date). Onrario: Thunder Bay District: Fort Williams (78, 22) (25 June); Orient Bay (113, 49) (22 June); Silver Inlet (26,19) (12-19 July). Nipissing District: Go Home Bay (18) (19 June); Cochrane (26) (no date); North Bay (2¢ ) (no date). QurBec: Angenteuill, Newago (2¢ ) (no date). Libellula (Ladona) deplanata Rambur Libellula deplanata Rambur, 1842, Histoire Naturelle des Insectes, Neurop- teres, pp. 75-76; Hagen, 1861, Synopsis of Neuroptera of North America, p. 154; Hagen, 1875, Proc. Boston Soc. Nat. Hist., vol. 18, p. 69; Kirby, 1890, Synoptic Catalogue of Neuroptera-Odonata, p, 26; Calvert, 1893, Trans. American Ent. Soc., vol. 208, p. 258; Williamson, 1900, Rpt. Indiana Dept. Geol., vol. 24, p. 331; Osborn, 1905, Ent. News, vol. 16, p. 195; Muttkowski, 1908, Bull. Wisconsin Soc. Nat. Hist., vol. 6, p. 115; Calvert, 1912, Ent. News, vol. 23, p. 484; Weiss, 1915, Ent. News, vol. 26, p. 102; Byers, 1927, Ent. News, vol. 38, p. 114; Byers, 1928, Ent. News, vol. 39, p. 230; Byers, 1930, Univ. Florida Pub. Biol. Sci., vol. 1, pp. 18, 26, 38, 107, 110-112, 117, 121, 232, 261, 275, 299; Williamson, 1931, Ent. News, vol. 42, p. 149; Montgomery, 1940, Jour. Mitchell Soc., vol. 56, p. 292. Ladona deplanata; Needham, 1897, Canadian Ent., vol. 29, pp. 141-146; Need- ham, 1901, Bull. New York St. Mus., vol. 47, pp. 528-530; Brimley, 1908, Ent, News, vol. 17, p. 185; Brimley, 1918, Ent. News, vol. 29, p. 229; Needham and Heywood, 1929, Handbook of North American Dragonflies, p. 217; Brimley, 1938, Insects of North Carolina, p. 39; Montgomery, 1940, Jour. Mitchell Soc., vol. 56, p. 292; Bick, 1950, American Midl. Nat., vol. 43, p. 73; Westfall, 1953, Florida Ent., vol. 36, p. 171; Montgomery, 1955, Proc. Indiana Acad. Sci., vol. 64, p. 184; Needham and Westfall, 1955, Dragonflies of North America, pp. 58, 474, 476, 477. Libellula (Ladona) deplanata; Borror, 1945, Ann. Ent. Soc. Amer., vol. 38, p. 192; Cook, 1951, Ent. News, vol. 62, p. 187; Kennedy, 1922, Ent. News, vol. 38, pp. 69, 110-111; Ris, 1910, Coll. de Selys, vol. 11, p. 258. Types. Male holotype and female allotype in the Selys Collec- tion, Institut Royal d'Histoire Naturelle de Belgique, Brussels, Belgium. Holotype with thorax damaged dorsally, left forewing broken and lost at level of second antenodal crossvein and proximal side of triangle; shows signs of repair on thorax, glue on legs and one spot on wings. Allotype with much of thorax gone on left side, all legs gone, right forewing broken and lost at nodus. A TAxONOMIC STUDY OF THE SUBGENUS LADONA 373 Mate. Measurements: Total length, 33.0 + 1.469 mm. (holotype 30.0 mm.); abdomen length, 21.6 + 1.608 mm. (holotype 18.0 mm., appendages missing); forewing length, 24.10 + 1.742 mm. (holo- type 25.5 mm.); hind wing length, 23.70 + 1.625 mm. (holotype 25.0 mm.); stigma length, 3.10 + .622 mm. (holotype 3.0 mm.); length nodus to stigma, 8.95 + .267 mm. (holotype 9.0 mm.); meta- thoracic femur length, 5.22 + .718 mm. (holotype 4.0 mm.); pro- thoracic tibia length, 3.45 + .444 mm. (holotype 4.0 mm.); meso- thoracic tibia length, 402+ .486 mm. (holotype 4.0 mm.); metathoracic tibia length, 5.03 = .179 mm. (holotype 4.0 mm.); head width, 5.96 + .720 mm. (holotype 5.5 mm.); superior anal appendage length, 1.68 + .276 mm.; supra-anal plate length, 1.24 + 377 mm. Head with parietal region of cranium brown; compound eyes tan or brown; vertex black with punctate appearance; ocelli sur- rounded by yellow margin; frons uniformly light tan, darkening with age; epistomal suture a colorless groove arched somewhat dorsally; postclypeus uniform light tan like frons; anteclypeus slightly darker than frons; labrum yellow, becoming slightly darker toward median depression; labium tan to yellow, median lobe darkening with age. Prothorax with dorsal margin of anterior lobe gold, remainder brown; middle lobe tan to light brown with median ridge bearing two dorsolateral dark brown spots; posterior lobe light tan to brown, occasionally darkening with age; dorsal lobe of eucervicale gold with occasional dark patches posteriorly. Synthorax with acrotergite II, mesostigmal plate, area posterior to pit of tergal apophysis II, and collar carina black; dorsal carina light brown; antealar carina light brown; antealar sinus light tan or brown; anepisternum II dorsally with mesal brown region bordered on each side by broad cream antehumeral stripe, darken- ing with age to blue-black; anepisternum II laterally with upper half cream, lower half dark brown; katepisternum II uniform light brown; mesopleural suture forming deep groove dorsally; epimeron II uniform light brown; anepisternum III tan; epimeron III light brown; metapleural suture dorsally a deep groove surrounded by dark brown area, remainder paler. Forewings with subcostal-radial and medial interspace dark brown to level of second antenodal crossvein; radial and medial- postcubital interspace without coloration; postcubital-anal inter- space with faint and irregular light brown extending approximately 374 THE UNIVERSITY SCIENCE BULLETIN half way to anal crossing; stigma light brown, subtending two to three cells; antenodal crossveins 10 to 12; postnodal crossveins 8 to 11. Hind wings with subcostal-radial and medial interspace dark brown to level of fourth or fifth antenodal crossvein; radial and medial-postcubital interspace without coloration; postcubital- anal interspace dark brown to middle of triangle; first anal inter- space with small irregular dark brown triangle; stigma light brown, subtending two to three cells; antenodal crossveins 9 to 10; post- nodal crossveins 9 to 11 (Fig. 9). Legs with femora and tibiae uniformly light tan to brown, bear- ing small dark brown spots at distal ends of femora; tarsi uniformly dark brown to black. Abdomen with segment I dark brown dorsally, light tan laterally and ventrally; segment II dorsally tan to light brown with anterior band of dark brown and with a dorsal dark brown triangular spot, laterally and ventrally tan to light brown; segment III dorsally tan to light brown with dorsal mesal light brown stripe becoming wide and black posteriorly, laterally tan with dorsal margin dark, becoming black and wider posteriorly, ventrally light tan; segments IV-VIII uniform with dorsal black stripe becoming wider posteri- orly in each segment to form a black triangle, laterally with black dorsal region becoming wider posteriorly, with remainder of lateral surface tan bearing black on ventral margin; segment IX dorsally uniform black, laterally tan without black marking on ventral margin, ventrally tan; segment X dorsally light brown with irregu- lar mesal black area, laterally light brown, ventrally tan; posterior ventral margin straight; supplementary transverse carinae shiny black, sinuate; posterior transverse carinae shiny black; submarginal transverse carinae shiny black; ventral carinae tan, occasionally becoming darker with age; acrotergites gray. Superior anal appendage brown to black; inferior anal lamina not widely separated. Penis with stem constricted, posterior projection sharply pointed; posterior margin of middle segment angular; distal segment a somewhat narrow pedicel; cornula sharply pointed (Fig. 4). Posterior hamulus bifurcated with anterior process bearing a single laterally directed hook; lateral lobe broadly rounded (Fig. 5). FEMALE. Measurements: Total length, 31.8 + 1.894 mm. (allo- type 30.0 mm. ); abdomen length, 19.7 + 1.728 mm. (allotype 18.5 + mm.); forewing length, 24.0+ 1.413 mm. (allotype 26.5 mm.); hind wing length, 23.0 + 1.412 mm. (allotype 24.0 mm.); stigma A Taxonomic StupDy OF THE SUBGENUS LADONA 375 length, 3.33 + .506 mm. (allotype 3.0 mm.); length nodus to stigma, 8.65 + .332 mm. (allotype 9.0 mm.); metathoracic femur length, 5.05 + .764 mm.; prothoracic tibia length, 3.43 + .636 mm.; meso- thoracic tibia length, 3.99 + .323 mm.; metathoracic tibia length, 5.01 + .366 mm.; head width, 5.87 + .383 mm. (allotype 5.0 mm.) superior anal appendage length, .76 + .171 mm. Head with parietal region of cranium light brown; compound eyes as male; vertex shiny black; ocelli with or without light yellow margin; frons yellow to orange; epistomal suture a slightly arched groove; postclypeus light brown with occasional yellow markings on ventral margin; anteclypeus as postclypeus; labrum yellow with irregular brown markings mesally; labium medium yellow with irregular brown markings on lateral portion of lateral lobes, darkening with age. Prothorax with dorsal margin of dorsal lobe gold, remainder black; middle lobe tan to yellow with irregular black lateral mark- ings; posterior lobe light brown with blue-black lateral markings; dorsal lobe of eucervicale yellow becoming brown posteriorly. Synthorax with acrotergite II and mesostigmal plate blue-black; area posterior to pit of tergal apophysis II brown to black; collar carina brown, dorsal carina yellow; antealar carina light tan be- coming black posteriorly; antealar sinus brown; anepisternum II dorsally and laterally as in male; katepisternum II greenish gray to olive; mesopleural suture forming deep groove dorsally and ventrally, surrounded by darkened areas; epimeron II greenish gray to olive; interpleural suture as a faint line; anepisternum III and katepisternum III greenish gray to olive; epimeron III greenish gray to olive with posterior traces of brown; metapleural suture deepened to form groove dorsally and ventrally. Coloration of lateral synthorax may become obscured with age due to irregular patches of blue-black coloration. Forewing with costal-subcostal interspace clear; subcostal-radial and medial interspace dark brown to level of second antenodal crossvein; radial and medial-postcubital interspace clear; post- cubital-anal interspace with very light translucent brown extending about half way to anal crossing; stigma tan, subtending two to three cells; antenodal crossveins 11 to 13; postnodal crossveins 8 to 10. Hind wing with costal-subcostal interspace clear; subcostal- radial and medial interspace dark brown to level of second or > third antenodal crossvein; radial and medial-postcubital interspace clear; postcubital-anal interspace dark brown to triangle; first anal 376 THe UNIvERSITY SCIENCE BULLETIN interspace with very small black triangle covering three to four cells; stigma tan, subtending two to three cells; antenodal cross- veins 9 to 10; postnodal crossveins 8 to 10. Legs with coxae and femora dark brown or tan with black areas at distal ends of femora; tibiae light tan; tarsi dark brown to black. Abdomen with segment I dorsally black, laterally and ventrally tan; segments II-X as in male. In both males and females of deplanata, there is occasionally a general darkening of the abdo- men with age. Anal appendage brown to black, elongate with broadly rounded apex, shorter than segment X; supra-anal plate tan to brown, more sharply rounded at apex than in julia, shorter than anal appendage. Nympu. Total length, 18-20 mm.; mentum with median projec- tion less prominent than in julia; no mental setae present; dorsal abdominal hooks overlapping half or more of following segment; length of anal appendages 3.0 mm., of abdominal segments [X-X 2.0 mm. Bronomics. Needham and Heywood (1929) indicated that this species is found flying about the edges or slightly over the open water of ponds. They also reported this species as having a slow, irregular flight and being easy to capture. Montgomery (1953) observed this species to fly with fluttering motions over bare areas along the shore, alighting on bare ground in a manner reminiscent of Plathemis lydia. However, Montgomery points out that de- planata is less active than lydia and does not fly rapidly along the lake margins as does lydia. Bick (1957) observed this species flying swiftly, approximately three feet above the water. He also observed deplanata perched on the ground, not more than twenty feet from the water’s edge. He states that this species occurs in pinelands, near lentic waters including sloughs, ponds, and borrow pits. Bick (1950) records having taken nymphs of this species from mud bottomed roadside borrow pits with heavy vegetation. He found the nymphs of deplanata associated with Tetragoneuria cynosura, Libellula lydia, Libellula luctuosa, Pachidiplax longi- pennis, and Erythemis simplicicollis. The more northern specimens of this species show a considerable darkening of most features of the color pattern. The synthorax tends to become blue-black and the antehumeral stripes less readily visible. The southern specimens were found to agree more with the type than those in the northern portion of the range. Due to these differences between the specimens from Florida and more A Taxonomic STUDY OF THE SUBGENUS LADONA Onl northern localities, there is some justification for subspecific rank for the more northern forms. However, due to the paucity of northern specimens and absence of material from intervening regions, I do not feel it advisable to attach formal names to the geographical variants at the present time. DistripuTion. Type locality: “North America.” ALABAMA: Baldwin Co. (14,12) (11 April). Frorma: Alachua Co. (104, 219) (15 March-3 April); Brevard Co. (1 ¢ ) (2 March); Lee Co. (14, 1¢) (21 March-3 April); Gadsden Co. (1?) (8 April); Gulf Co, (48, 82) (25 March-12 April); Highlands Co. (8¢, 13 ¢ ) (12 Feb.-20 March); Hillsboro Co. (22 ) (12 Feb.-8 March); Leon Co. (54, 52) (28 March); Madison Co. (26, 292) (6 April); Okeechobee Co. (14,22) (18 Feb.); Orange Co. (204, 172) (1 Jan.-10 Dec.); Osceola Co. (26, 142) (23 Feb.-22 March); Palm Beach Co. (1¢ ) (24 Feb.); Pinellas Co. (146, 29 ) (4 April); Sarasota Co. (29 ) (—Feb.-1 April); Seminole Co. (114, 82) (14 Jan.-5 March); Suwannee Co. (12 ) (28 April); Volusia Co. (54,42 ) (22 April); Wakulla Co. (24 ) (1 April). Inprana: Clark Co. (124, 31¢) (11-12 June). Lovutstana: Union Parish (14 ) (1 April); St. Tammany Parish (12 ) (9 April). M1ussissrppr: George Co. (16, 192) (10-26 April); Harrison Co. (16) (10 April). Norra Carouina: Craven Co. (22?) (14 May). Oxta- HOMA: Pushmataha Co. (1 ¢ ) (28 March). Libellula (Ladona) exusta Say Libellula exusta Say, 1839, Jour. Acad. Nat. Sci., Philadelphia, vol. 8, pp. 29-30; Hagen, 1861, Synopsis of Neuroptera of North America, p. 115; Hagen, 1873, Proc. Boston Soc. Nat. Hist., vol. 16, p. 361; Kirby, 1890, Synoptic Catalogue of Neuroptera-Odonata, p. 26; Wadsworth, 1890, Ent. News, vol. 1, p. 56; Harvey, 1891, Ent. News, vol. 2, p. 4; Hitchings, 1892, Ent. News, vol. 3, p. 39; Calvert, 1893, Trans. American Ent. Soc., vol. 20, pp. 152, 258; Calvert, 1895, Jour. New York Ent. Soc., vol. 3, p. 47; God- dard, 1896, Proc. American Philos. Soc., vol. 35, p. 209; Calvert, 1897, Jour. New York Ent. Soc., vol. 5, p. 47; Harvey, 1898, Ent. News, vol. 9, p. 87; Hine, 1899, Ann. Rpt. Ohio Acad. Sci., vol. 7, p. 66; Kellicott, 1899, Special Pub. Ohio Acad. Sci., vol. 2, p. 99; Burnham, 1900, Proc. Man- chester (N. H.) Inst. Art. Sci., vol. 1, p. 35; Osburn and Hine, 1900, Ohio Nat., vol. 1, p. 14; Weith, 1900, Ent. News, vol. 11, p. 641; Wil- liamson, 1900, Rpt. Indiana Dept. Geol., vol. 24, pp. 327, 328, 331; Williamson, 1900, Proc. Indiana Acad. Sci., vol. 24, p. 173; Calvert, 1900, Rpt. New Jersey Board Agric., vol. 27, p. 8; Bumham, 1901, Proc. Man- chester (N.H.) Inst. Art. Sci., vol. 2, p. 44; Harvey, 1902, Ent. News, vol. 18, p. 10; Brimley and Sherman, 1904, Ent. News, vol. 15, p. 101; Calvert, 1905, Occ. Pub. Boston Soc. Nat. Hist., vol. 7, p. 34; Osborn, 1905, Ent. News, vol. 16, pp. 186, 195, 196; Williamson, 1907, Ohio Nat., vol. 7, p. 105; Muttkowski, 1908, Bull. Wisconsin Nat. Hist. Soc., vol. 6, p. 116; Walker, 1908, Ottawa Nat., vol. 22, p. 23; Wilson, 1909, Proc. United States Nat. Mus., vol. 36, pp. 655, 657, 658, 670; Calvert, 1910, Rpt. New Jersey State Mus., 1909, p. 80; Davis, 1913, Jour. New York 15a—1367 378 THE UNIVERSITY SCIENCE BULLETIN Ent. Soc., vol. 21, p. 24; Whedon, 1914, Rpt. Minnesota St. Ent., vol. 15, p. 98; Woodruff, 1914, Jour. New York Ent. Soc., vol. 22, p. 158; Walker, 1915, Rpt. Ontario Dept. Fish., vol. 47 (supplement), pp. 58, 60; Hankin- son, 1916, Pub. Michigan Geol. Sur., vol. 20, p. 117; Howe, 1916, Univ. Michigan Mus. Zool., Occ. Pap., no. 41, p. 8; Howe, 1917, Psyche, vol. 24, p. 51; Williamson, 1917, Misc. Pub. Univ. Michigan Mus. Zool., vol. 2, p. 10; Howe, 1919, Canadian Ent., vol. 51, p. 14; Howe, 1919, Ent. News, vol. 30, p. 13; Howe, 1919, Psyche, vol. 26, p. 67; Howe, 1920, Psyche, vol. 27, p. 57; Howe, 1920, Mem. Thoreau Mus., vol. 2, pp. 67, 71, 97, 101; Howe, 1921, Proc. Boston Soc. Nat. Hist., vol. 36, pp. 114, 115, 128; Howe, 1921, Mem. Thoreau Mus., vol. 2 (supplement), p. 10; Williamson, 1921, Ent. News, vol. 32, p. 22; Byers, 1925, Pub. Michigan Acad. Sci., vol. 5, pp. 393, 396; Byers, 1927, Ent. News, vol. 38, p. 114; Byers, 1927, Univ. Michigan Mus. Zool., Occ. Pap., no. 183, p. 11; Kennedy, 1928, Ecology, vol. 9, p. 375; Byers, 1930, Univ. Florida Pub. Biol. Sci., vol. 1, pp. 111-112; Walker, 1934, Proc. Nova Scotia Inst. Sci., vol. 18, p. 126; Borror, 1940, Ent. News, vol. 51, p. 78; Walker, 1942, Proc. Nova Scotia Inst. Sci., vol. 20, p. 174. Ladona exusta; Needham, 1897, Canadian Ent., vol. 29, pp. 145-146; Needham, 1901, Bull. New York St. Mus., vol. 47, p. 528; Needham, 1908, Rpt. Michigan Geol. Sur., 1907, p. 264; Muttkowski, 1910, Bull. Wisconsin Nat. Hist. Soc., vol. 8, p. 57; Needham and Heywood, 1929, Handbook of Dragonflies of North America, pp. 217, 218, 226; Borror, 1935, Ohio Jour. Sci., vol. 35, p. 454; Wright, 1937, Jour. Tennessee Acad. Sci., vol. 12, p. 260; Wright, 1939, Jour. Tennessee Acad. Sci., vol. 14, p. 207; Wright, 1943, Jour. Tennessee Acad. Sci., vol. 18, p. 189; Needham and Westfall, 1955, Dragonflies of North America, pp. 58, 474, 476, 477; Kormondy, 1959, Misc. Pub. Univ. Michigan Mus. Zool., no. 104, p. 34. Libellula (Ladona) exusta; Borror, 1945, Ann. Ent. Soc. Amer., vol. 38, p. 192; Kennedy, 1922, Ent. News, vol. 33, p. 111; Ris, 1910, Coll. de Selys, vol. 11, pp. 247, 256-259. Types. Male holotype located in Museum of Comparative Zool- ogy, Harvard University, Cambridge, Massachusetts. Mar. Measurements: Total length, 36.8 + 1.894 mm. (holo- type 39.4 mm.); abdomen length, 23.4 + 1.435 mm. (holotype 24.8 mm.); forewing length, 27.8+1.316 mm. (holotype 33.0 mm.); hind wing length, 26.8 + 1.316 mm. (holotype 31.2 mm.); stigma length, 2.91 + .423 mm. (holotype 3.2 mm.); length nodus_ to stigma, 10.04 + .610 mm. (holotype 11.2 mm.); metathoracic femur, 5.75 + 1.00 mm. (holotype 8.0 mm.); prothoracic tibia length, 3.68 + 806 mm.; mesothoracic tibia length, 4.21 + .655 mm.; meta- thoracic tibia length, 5.62 + .806 mm. (holotype 6.6 mm.); head width, 6.15 + .948 mm. (holotype 6.7 mm.); superior anal append- age length, 1.57 + .236 mm. (holotype about 2.0 mm.); supra-anal appendage length, 1.13 + .212 mm. (holotype about 1.0 mm.). Head with parietal region of cranium light brown, darkening with age; compound eyes reddish brown; vertex with dorsal two- thirds yellowish tan to olive, ventral one-third brown; ocelli sur- rounded by yellow band; antenna black; frons light tan to olive, darkening laterally; epistomal suture an arched groove; postclypeus A TaxONOMiIc STUDY OF THE SUBGENUS LADONA 379 tan to dark brown; anteclypeus like frons in color; labrum tan to dark brown; labium colored like labrum, darkening to black with age. Prothorax with anterior lobe light yellow to gold, bearing irregu- lar central brown markings; middle lobe light brown becoming darker mesally, median ridge brown with white hairs; postericr lobe brown becoming darker posteriorly, sometimes becoming black with age; dorsal lobe of eucervicale uniform gold to yellow, darkening with age. Synthorax with acrotergite II black becoming light tan laterally; mesostigmal plates yellow to light brown; area posterior to pit of tergal apophysis II tan to light brown with irregular mesal mark- ings of darker brown; collar carina light brown; dorsal carina blue-black; antealar carina tan becoming shiny black posteriorly; antealar sinus tan; anepisternum II dorsally brown bordered by antehumeral stripes of yellow-gray, darkening to black with age; anepisternum II laterally with upper half yellow-gray to cream, lower half dark brown; katepisternum II brown, darkening with age; mesopleural suture deepened dorsally and ventrally; epimeron II brown; anepisternum III brown; katepisternum III brown; epi- meron III brown; metapleural suture deepened dorsally and ven- trally. Forewing with subcostal-radial and medial interspace dark brown to level of fourth antenodal crossvein; radial and medial-postcubital interspace light translucent brown to arculus; postcubital-anal inter- space dark brown to anal crossing or slightly beyond; stigma light tan, subtending two to three cells; antenodal crossveins 11 to 14; postnodal crossveins 9 to 18. Hind wing with subcostal-radial and medial interspace dark brown to level of fifth or sixth ante- nodal crossvein; radial and medial-postcubital interspace light translucent brown to level of fifth or sixth antenodal crossvein; postcubital-anal interspace dark brown to anal crossing or slightly beyond; first anal interspace with dark brown triangle; stigma tan, subtending two to three cells; antenodal crossveins 9 to 12; post- nodal crossveins 10 to 12 (Fig. 10). Legs with femora and tibiae undifferentiated light tan, becoming brown to black with age; tarsi brown. Abdomen with segment I dorsally olive to light tan, laterally and ventrally light brown; segments H-IX in young males dorsally and laterally dull white with blue-black traces, in mature males black; all carinae tan except black dorsal and posterior transverse 380 THe UNIVERSITY SCIENCE BULLETIN carinae, ventral aspect of segments tan, these segments darkening to black with occasional irregular pale areas remaining, all carinae with exception of ventral carina becoming shiny black; segment X light tan with concave ventral posterior margin. Superior anal appendage brown to black, darkening with age; inferior anal lamina not widely separated. Penis without constricted stem, posterior projection broadly rounded; posterior margin of middle segment angular; distal seg- ment a broad pedicel; cornula broadly rounded at each end (Fig. 6). Posterior hamuli bifurcate with anterior process bearing two laterally directed hooks; lateral lobe bluntly rounded (Fig. 7). FemMate. Measurements: Total length, 33.2 + 1.447 mm.; ab- domen length, 20.8 + 1.707 mm.; hind wing length 24.8 + 1.600 mm.; forewing length, 25.8 + 1.927 mm.; stigma length, 3.33 + .547 mm.; length nodus to stigma, 9.76 + .387 mm.; metathoracic femur length, 5.63 + .788 mm.; prothoracic tibia length, 3.64 + .572 mm.; mesothoracic tibia length, 4.05 + .526 mm.; metathoracic tibia length, 5.59 + .726 mm.; head width, 5.91 + .924 mm.; superior anal appendage length, .85 = .179 mm. Head with parietal region tan; compound eyes reddish brown; vertex with dorsal two-thirds reddish brown, ventral one-third black; ocelli surrounded by yellow band; frons tan to olive; epis- tomal suture an arched groove; postclypeus as frons; anteclypeus as frons; labrum yellowish, darkening mesally; labium colored like labrum with occasional irregular black markings. Prothorax with anterior lobe black bearing lateral yellow spots; middle lobe brown becoming darker mesally; posterior lobe light brown to black; dorsal lobe of eucervicale gold. Synthorax with acrotergite Il black becoming light brown lat- erally; mesostigmal plate tan, area posterior to pit of tergal apophy- sis II black becoming light brown laterally; collar carina tan; dorsal carina black; antealar carina anteriorly brown becoming black posteriorly; antealar sinus tan; anepisternum II dorsally light brown, bordered laterally by gray antehumeral stripe; anepis- ternum II laterally with upper one-third gray, middle one-third dark brown, remainder medium brown; katepisternum II medium brown; mesopleural suture deepened dorsally and ventrally form- ing deep grooves surrounded by blue-black areas; epimeron II, anepisternum III, katepisternum III, epimeron HI medium brown A TaxoNomMic STUDY OF THE SUBGENUS LADONA 881 as katepisternum II, showing irregular darkening with age; inter- pleural suture as a faint groove; metapleural suture deepened dorsally and ventrally, markedly so ventrally. Forewing with costal-subcostal interspace with irregular trans- lucent brown to level of second antenodal crossvein; subcostal- radial and medial interspace dark brown to level of fourth or fifth antenodal crossvein; radial and medial-postcubital interspace trans- lucent brown to arculus or slightly beyond; postcubital-anal inter- space dark brown to level of arculus or slightly beyond; stigma yellow to light tan, subtending two to three cells; antenodal cross- veins 10 to 14; postnodal crossveins 8 to 10. Hind wing with costal-subcostal interspace as in forewing; subcostal-radial and medial interspace dark brown to level of sixth or seventh antenodal crossvein; radial and medial-postcubital interspace translucent brown to level of sixth or seventh antenodal crossvein; postcubital- anal interspace dark brown to middle of triangle; first anal inter- space with dark brown triangle; stigma as forewing, subtending two to three cells; antenodal crossveins 9 to 11; postnodal cross- veins § to 12. Legs with femora light brown; tibiae light tan to light brown; tarsi brown, slightly darker than tibia. Abdomen with segment I dorsally black becoming brown laterally and ventrally; segment II dorsally and laterally dark brown with traces of black, ventrally light tan to light brown; segment III dorsally and laterally dark dull brown with irregular black, ven- trally light brown; segments IV-IX with regular black stripe dor- sally, laterally dark dull brown with dorsal and ventral black stripes, occasional irregular black, ventrally dark brown; segment X dark, almost black dorsally, brown ventrally; supplementary carinae shiny tan; posterior transverse carinae brown; submarginal ventral carinae black; ventral carinae brown; acrotergites and ante- costal sutures brown to black. Anal appendages as in deplanata. NympuH. Unknown. Bronomics. Evidently no biological data in literature. DistRiBuTION. Type locality: “Massachusetts.” CONNECTICUT: Litchfield Co. (104, 22) (20 June). Mare: Androscoggin Co. (39) (21 June); Aroostock Co. (32 ) (30 May-18 June); Kenne- bee Co. (58, 72) (11 Jume-5 July); Lincoln Co. (22¢, 2892) (9 June-28 July); Penobscot Co. (14 ) (12 July); Piscataquis Co. 382 Tue UNIVERSITY SCIENCE BULLETIN (39) (6 April). New Hampsuire: Carroll Co. (12 ) (28 July). New Jersey: Burlington Co. (14,39 ) (6-10 June); Camden Co. (1g, 12) (20 May); Middlesex Co. (1?) (5 June); Ocean Co. (26, 162) (2 May-3 July). Massacuuserts: Barnstable Co. (114) (10 June-1 July); Hampden Co. (14 ) (20 June); Middle- sex Co. (12) (21 June); Worcester Co. (16) (18 June). New York: Essex Co. (29) (4 July). Vermont: Windham Co. (2 ¢ ) (30 April). In addition to the above material, the following specimens were examined from Canada. Nova Scorra: Hants, Mt. Uniacke (2¢, 32) (20 June). VARIATIONS IN LIBELLULA (LADONA) JULIA With a few notable exceptions, the taxonomy of Odonata has progressed little beyond the alpha level because the field as a whole has been slow in accepting modern systematic methods. For this reason, the species limits are somewhat poorly defined, and infraspecific variation is rarely considered. U2 lila LOS5mm. cise FIG.6 Tue UNIVERSITY SCIENCE BULLETIN ‘pysnxa “J ‘OT ‘Sty ‘pypunjdap “7 “6 “St yo uoriod [eseq ul WoTeAO[OO govds1ayUL JO JUa}X9 SoywoIpUL voIe PousyxIep -DUOpVT JO satoads UT UORIO[OD SUIAA ‘“OT-§ ‘SOI 6 9l3s 8°9l4 aN ES N \) ayy as } CELL A TAXONOMIC STUDY OF THE SUBGENUS LADONA 393 5.80 5.50 5.20 4.90 yi Aa | | | / Be 4.30 | | Ea ma 4.00 | | 3.70 POP | 2 3 4 5 3.60 Fic. 11. Comparison of length of front tibia in five populations of males of Libellula (aaeae julia. Symbols: mean (dot), standard error of mean (cross bars), three standard deviations either side of mean (vertical line), and observed range of variation (rectangle). Measurements in millimeters. 394 Tue UNIVERSITY SCIENCE BULLETIN 7.90 7.60 030 LF | x00 li i em a is ‘aad il 6.30 POP | 3 4 ) 6.00 Fic. 12. Comparison of head width in five populations of males of Libellula (Ladona) julia. Symbols as in Fig. 11. A TAXONOMIC STUDY OF THE SUBGENUS LADONA 7.50 7.20 6.90 5.70 5.40 POP | 5.10 + 2 395 Fic. 13. Comparison of length of metathoracic femur in five populations of males of Libellula (Ladona) julia. Symbols as in Fig. 11. 396 THe UNIVERSITY SCIENCE BULLETIN 3.90 3.50 piston one | | CSE eens Eee | eee 3.20 HIE ] no 2.99 2.96 POP | 2 3 4 5 Fic. 14. Comparison of length of stigma in five populations of males of Libellula (Ladona) julia. Symbols as in Fig. 11. THE UNIVERSITY OF KANSAS SCLENCE BULLETIN VoL. XLV] JuNE 7, 1965 [No. 5 The genus Ptilomera Amyot and Serville (Gerridae: Hemiptera) * BY HERBERT B. HUNGERFORD and Ryuricur MaAtsupa ** AgBsTRACT: This paper redescribes and illustrates all types and gives a key to both males and females. Resurrects from synonymy: P. laelaps Breddin, P. oribasus Breddin and P. swmizome Esaki. Accepts P. asbolus Breddin and P. argus Breddin as synonyms of P. dromas Breddin. Makes P. harpyia Schmidt a synonym of P. tigrina Uhler with a reser- vation, P. lachne Schmidt as the synonym of P. agriodes Schmidt, P. canace Schmidt as synonym of P. cingalensis Stal. Describes thirteen species as new: P. assamensis from Assam; P. chinai, P. nunikanensis, P. kirkaldyi, P. sarawakensis, P. maai, P. gressitti from Borneo; P. lundbladi, P. sumatranus from Sumatra; P. swmbaensis from Sumba Island; P. timorensis from Timor Island; P. breddini, P. papuensis from New Guinea. The paper also describes P. harpyia ceramensis as a new subspecies from Ceram, an island in the Molluccas and P. aéllo cheesmanae as a new subspecies from New Guinea. The genus Ptilomera Amyot and Serville 1843. Amyot et Serville, Hist. nat. Ins. Hémipt. I:413 (for Gerris laticauda Hardw. (They misspelled G. laticaudata Hardw. ) 1853. Herrich-Schaeffer, Wanz. Insekten 9:65, Taf. 305, fig. 940. (Recorded and figured a species from Java which was not P. laticaudata. ) 1901. Breddin, Abh. Naturf. Ges. Halle, 24:86-89 Taf. 1, figs. 10-14. (De- scribed P. pamphagus, P. dorceus, P. oribasus, P. laelaps é .) 1903. Distant, Faun. Brit. Ind. Rhynch. 2:185. 1903. Breddin, Societas Entomologica 17:147-48. (Described four new species, P. laelaps 2, P. dromas, P. argus, P. hylactor. ) * Contribution number 1160 from the Department of Entomology, The University of Kansas. This study was made possible with the aid of a grant from the National Science Foundation. ** The senior author takes primary responsibility for the recognition of species and for the keys and descriptions of species and the junior author for the measurements and illustrations. However, we have checked each other’s work throughout. (397 ) 16—1367 398 THE UNIVERSITY SCIENCE BULLETIN 1905. Breddin, Mit. Naturh. Mus. Hamburg, 22:132-135. (Described P. asbolus, P. argus.) 1906. Breddin, Societas Entomologica, 21(2):9. (Described P. aéllo.) 1909. Kirkaldy, Canad. Ent. 41:390. (Described Ptilomera (Rheumatogonus ) subg. new. ) 1925. Esaki, Philip. Jour. Sci. 26(1):59-60, Pl. 1, figs. 7-8. 1926. Schmidt, Ent. Mitteil. 15(1):63. (Described five new species, P. agriodes, P. lachne, P. canace, P. harpyia, P. harpalos.) 1927. Esaki, Rev. Esp. Ent. (3):251, 254, 258-261. (Described Ptilomerinae as new subf., keyed out genera, listed seventeen species of Ptilomera.) 1933. Lundblad, Arch. Hydrobiol. Suppl. 12, Tropische Binnengewisser 4:415- 429, figs. 133-137, Taf. 13. (Proposed many synonyms. ) 1934. Lundblad, Ark. Zool. 27A(14):26. (Key to genera of Ptilomerinae. ) 1941. Kenaga, Univ. of Kansas Sci. Bull. 27(1):169. 1942. Kuitert, Univ. of Kansas Sci. Bull. 28(1):118. 1960. Hungerford and Matsuda, Univ. of Kansas Sci. Bull. 41(1):7. (Key to genera and subgenera of Ptilomerinae. ) 1960. Matsuda, Univ. of Kansas Sci. Bull. 41(2):261-270, figs. 33, 36, 119- 120, 137, 157, 166, 609-641. (Description of genus Ptilomera.) Ptilomera Amyot and Serville belongs to the gerrid subfamily Ptilomerinae and can be separated from the other seven genera of this subfamily by our “Key to subfamilies, tribes, genera and sub- genera of the Gerridae of the World” [Univ. of Kansas Sci. Bull. 41(1):4-5, 7-8, (1960) ]. This genus is easy to recognize because all the species are rather large and stout bodied waterstriders with marked sexual dimorphism. The males have a striking fringe of hairs on the rear margin of the distal half, at least, of the middle femur. This is lacking in the female. The male genital segments are large and conspicuous while those of female are not. In this genus the males are often larger than the females which is unusual in the Gerridae. The anterior margin of the head is not bentventrad, widened posteriorly between eyes which are small, broadly rounded on the outer margin and quite broadly concave on inner margin. Antenniferous tubercles divergent anteriorly. Antennae shorter than the body. First segment longer than three following segments to- gether; third segment longer than either second or fourth; fourth segment curved in apical third. Beak short, not extending beyond prosternum. Apterous forms: The pronotum is subquadrangular in shape, about as wide as or a little wider than head including eyes, wider than long at middle, lateral margins broadly rounded, posterior margin nearly straight or feebly concave. Mesonotum about twice to two and a half times as long as pronotum, without either longi- tudinal sulcus or lateral longitudinal suture separating mesonotum from mesopleuron. Rear margin of mesonotum medially broadly Genus PrILOMERA AMYOT AND SERVILLE 399 prolonged and truncate to concave at rear margin. Metanotum with a median longitudinal depressed black line. Front leg with femur straight, a little longer than tibia; tibia with a conspicuous process at inner apical angle; apical inner surface of femur and basal inner surface of tibia with some tubercles; tarsus a little shorter than tibia, first tarsal segment about twice to a little less than three times as long as second; second segment slightly thickened near apex, claws arising from near apex and with a distinct arolium. Middle leg relatively longer in male than in female; femur of male robust, with a fringe of long hairs on rear margin, except in basal one-fourth; femur in female bare and slender; tibia more or less strongly curved, with shorter fringe of hairs on inner margin in both sexes; first tarsal segment about ten times as long as second segment; second segment with small claws arising from near apex. Hind leg longer than middle leg, relatively longer in male than in female; coxa with a small spine on apical margin except in the subgenus Proptilomera Hungerford and Matsuda; femur a little less than twice as long as tibia; both femur and tibia without fringe of hairs; first and second tarsal segments completely fused, very short, claws arising from near apex. Macropterous forms: Pronotum with anterior lobe well defined by transverse groove; humeri located at apical third of pronotum, posterior margin broadly rounded. Hemelytra with R + M and Cu veins basally distinct from each other, the two veins basally con- nected by a cross vein beyond middle of hemelytra. Vein A con- nected with Cu at about apical third of the wing. Hind wing with vein A distinct and connected with Cu at apical third of wing. Mesosternum has two longitudinal sutures that are not present in apterous forms. The color and color pattern are practically generically constant. Ground color of all but two species varies from pale testaceous to ferrugineous and the black markings occupy the same relative posi- tions in all, and are more or less covered with a silvery pubescence. Clypeus and antenniferous tubercles brown to black. Head with two pairs of black dots that may be connected into converging lines, and in dark forms also have a black line paralleling inner margin of eyes. Anterior margin of pronotum with a black spot, usually covered by silvery pile behind inner margin of each eye. These spots may or may not be joined. Anterolateral black spot is present on either side of mesonotum. Anterior and lateral margins at least of abdominal tergites dark brown to black. In side view, a brown 400 THE UNIVERSITY SCIENCE BULLETIN to black spot on each acetabulum, meso- and metapleural areas covered by a broad black longitudinal band, which may be divided into two bands either by a longitudinal stripe of slivery hairs or by a yellow stripe or by both, with silvery stripe on top of the yellow stripe. A dark band on sides of the abdomen below connexivum. Front femur with two dark longitudinal stripes, one above and one outside. Venter of thorax and abdomen from light brown to white and covered by a silvery pubescence. Distribution: Confined to the Oriental region and some islands be- yond it. In the Continental Asia it is known to occur in Hongkong, China, South India and Kashmir in the north, and eastward through Nepal, Upper Bengal and Assam, southward in Burma and Thailand and eastward in Indochina from Tonkin and Annam to Cambodia. It has been found in the following islands: Ceylon, Sumatra, Java, Borneo, Celebes, Timor, Ceram, New Guinea, Philippines and Formosa. The most primitive living species of Ptilomera we believe to be Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda from Singla Darjiling, E. Himalayas, and the species we believe to be farthest removed from the norm are Ptilomera (Ptilomera) werneri Hungerford and Matsuda from the Philippines and Ptilo- mera (Ptilomera) timorensis n. sp. from Timor Island. TAXONOMIC PROBLEMS INVOLVED When we undertook this study the problems involved in determin- ing species in this genus seemed to us almost unsurmountable. We had plenty of warning. Esaki (1927), who proposed the subfamily Ptilomerinae and gave a key to its genera, stated that the genus Ptilomera contained seven- teen described species and that “This genus is taxonomically one of the most difficult groups, as all the species are more or less varia- ble to some extent, and apparently much allied to one another.” He merely listed the species. Dr. Lundblad’s (1933) study of the genus Ptilomera has been indispensable to us. He had examined all the types except those of P. tigrina Uhler, P. dorceus Breddin and P. sumizome Esaki. It was his conclusion that this is a most difficult genus, and that the females appear to be best for specific determination but are hope- lessly variable as shown in his figs. 134, and 136 (see page 489). Secondly that the males appeared to him to be so variable that he did not use them in reaching his conclusion on synonymy. He re- Genus PritoMERA AMYOT AND SERVILLE 401 duced six species to synonymy and questioned two others. Thirdly, he thought that the various types might be races capable of inter- breeding, and that only the study of much material and experiments in rearing might settle the questions. How much material he studied besides the types we do not know. He recorded P. dromas Breddin from six collections in Java, only three of them had both sexes represented. A total of nineteen males and 35 females. Of P. pamphagus (his determination, not Bred- din’s) Lundblad recorded twelve collections from Sumatra only five of which had both sexes represented. A total of eleven males and thirteen females. With the above material it is understandable that he was perplexed. However, we have already described two new species and in this paper are describing thirteen more. We believe the male char- acters are quite stable and in no species have we found the females showing the hopeless variations suggested by Lundblad (1933) in his figures 134 and 136 (which we have reproduced on page 489). In fig. 136 Lundblad figured the female type of P. pamphagus Bred- din (A), the female type of P. oribasus Breddin (G), the female type of P. laelaps Breddin (1) all from Celebes, and six other draw- ings of females from Sumatra. He believed all nine drawings to be showing the variation in the female of P. pamphagus Breddin. When we found three males and three females of P. laelaps Breddin in our Kirkaldy Collection bearing the same label as the female type, we discovered that all three females were like the type of laelaps and the males were not like the type of P. pamphagus Breddin. We also found in the British Museum Collection a male and a female of P. oribasus Breddin. The female is like Breddin’s type and the male is not like either P. pamphagus Breddin or P. laelaps Breddin. Un- doubtedly, here are three good species from Celebes. We believe that the females of some species show some variations in the length of the connexival spines and in the shape of the caudal lobes of the seventh abdominal segment, but in other species such variations are difficult to find. In the males there may be a little variation in the shape of the pygofer or of the paramere but such variations within a species are small. This genus Ptilomera is, in our experience, almost unique. In many genera it is impossible to identify the females to species be- cause they all look so much alike while the males possess good specific characters. In the genus Ptilomera it is the complexity of form of the females that trouble us. 402 THe UNIVERSITY SCIENCE BULLETIN There were other difficulties. Hardwicke (1825) described Gerris laticaudata from Nepal and said it was eleven lines long. If his line was 1/12 inch then it should have been more than 23 mm. long. Lundblad (1933) found the type to be around 14mm. long. Hard- wicke had before him two specimens, a macropterous form and an apterous form which he thought were females and the apterous one anymph. They are both adult males. Yet this error is sex deter- mination was followed by Uhler (1860) when he described P. tigrina from three females and thought that they were males. It was not until Breddin (1901), were the sexes correctly recognized. The pioneer students of the genus Ptilomera, Breddin and Schmidt made too much of color in their descriptions. The color pattern is basically the same in all species with the black spots, covered by a silvery pile occupying the same positions in all species. The ground color varying from light to dark within a given species. Hard- wicke’s Fig. 4 was not correctly drawn, and Uhler did not give a figure. Breddin published figures of the apical abdominal segments of some of his species. The drawings of his female types are useful but those of the males are too small to show the real differences among the species he figured. Schmidt gave no drawings to help identify his species. While the above authors sometimes compared the size of one species with another, we have often found more variation in size within a species than between species. The former authors some- times gave length of the leg segments, especially the middle and hind femora, and described the fringe of hairs on the middle femur of the males. We have not found these characters to be of much use in specific determination. We began our studies by trying to measure all segments of the middle and hind legs but the tibia and tarsi are so often curved that it seemed impossible to measure ac- curately any of these save the femur. The earlier authors who described species from a single sex of either male or female gave us more names than species. For ex- ample, Schmidt described a male and female from South India as different species. When we began this study we had before us the following names. Synonyms suggested by Lundblad (1933) are shown in parentheses. 1825. P. laticaudata (Hardwicke), Nepal, male. 1855. P. cingalensis Stal, Ceylon, female. 1860. P. tigrina Uhler, Hongkong, China, female. 1901. P. dromas Breddin, Java. Drawing of female in 1901, description of male and female in 1903. East and South Java. Genus PTILOMERA AMYOT AND SERVILLE 403 1901. P. dorceus Breddin, Celebes. Described and figured female (same as P. dromas? ). WO, 12 porpliaeys Breddin, Celebes. Described and figured male and fe- male. 1901. P. oribasus Breddin, Celebes. Described and figured female (Syn. of _ P. pamphagus). 1901. P. laelaps Breddin, Celebes. Described and figured male, described fe- male in 1903 (syn. of P. pamphagus). 1903. P. hylactor Breddin, Annam. Described male and female. 1903. P. argus Breddin, W. Java. Described male (syn. of P. dromas). 1905. P. asbolus Breddin, W. Java. Described and figured male and female (syn. of P. dromas). 1906. P. aéllo Breddin, New Guinea. Described female. 1925. P. sumizome Esaki, Celebes. Described and figured male and female (syn. of P. pamphagus ). 1926. P. agriodes Schmidt, South India. Described male. 1926. : ee Schmidt, South India. Described female (syn. of P. agrio- es). 1926. P. canace Schmidt, Ceylon. Described male and female (syn. of P. cingalensis ). 1926. P. harpyia Schmidt, Cambodia. Described female. 1926. P. harpalos Schmidt, W. Sumatra. Described male and female. 1931. P. shirakii Esaki, Formosa (nomen nudum). 1958. P. werneri Hungerford and Matsuda, Philippines. Described and figured male and female. 1958. P. himalayensis Hungerford and Matsuda, Himalaya. Thus there were twenty one names, one of which is a nomen nudum, and eight possible synonyms as suggested by Dr. Lundblad. Moreover, only ten species were known from both sexes. Three species from males only and seven species from females only. From our preliminary studies of this genus we found the male characters quite stable and in no species did the females show the hopeless variation suggested by Lundblad (1933) in his figures 134 and 136. It seemed imperative that we re-examine the types of all species of Ptilomera. This we were permitted to do by the kindness of the following: Dr. R. Sailer of the U. S. National Museum in Washington D. C., Dr. W. E. China and Mr. R. J. Izzard of the British Museum in London, Dr. T. Jaczewski of the Polish Academy of Sciences in Warsaw, Dr. H. Sachtleben of the Deutsches En- tomologisches Institut in Berlin and Mr. T. Hidaka of the Kyushu University in Fukuoka, Japan. Our task has been to redescribe and illustrate all types, find females to match three male types and males to match seven female types and prove or disprove all synonymys. This we have been able to do for all but two. We need males of P. tigrina Uhler from Hongkong, China and of P. dorceus Breddin from Celebes. Although Lundblad saw the male type of P. pam- phagus from Celebes, he illustrated instead (his fig. 137, 1933) a 404 THe UNIVERSITY SCIENCE BULLETIN male specimen from Sumatra which is not P. pamphagus. Probably it is the male of the same species as the female he figured as B on his fig. 136. They came from the same place. We have found that the males provide a number of useful char- acters for the separation of species. The shape of the pygofer, the suranal plate and the paramere are remarkably constant for a species, so is the position of the metacoxae. 21 units in the measure- ments of relative lengths of various structures equal to 1mm. For the terminologies used in the description refer to pl. I. Genus PrmOoMERA AMYOT AND SERVILLE 405 JP bye IE dy IL Yeah ae Antenniferous tubercle ———— Pronotum _-Mesonotum _Metanotum /Metathoracic spiracle / / Za Za A ’ ~Metacetabula ea -—Mesoacetabula-~ Ieee Ist abd. spiracle ~~~ “~{ st abd. tergite < fon ~Metacoxa Sin abe: \eigiic = = ~ _elst gen. seg, _Lateral projection of pygofer _-~_-Lateral wing of suranal plate SS “ ~_-—Median lobe of suranal plate Connexival spine = Paramere =—-——Connexival spine Median ventral lobe~—_ _Nentrolateral lobe _-Dorsolateral lobe Fic. 1. Male of Ptilomera (Ptilomera) dello cheesmanae Hungerford and Matsuda. Fic. 2. Left paramere of Ptilomera (Ptilomera) sumbaensis Hungerford and Matsuda. Fic. 3. Female abdomen of Ptilomera (Ptilomera) harpalos Schmidt. Fic. 4. Lateral view of female seventh abdominal segment of Ptilomera ( Ptilomera) sumbaensis Hungerford and Matsuda. Fic. 5. Ventral view of female seventh abdominal segment of Ptilomera (Ptilomera) sumbaensis Hungerford and Matsuda. 406 2.(1) 4.(3) 5. (4) 6.(5) 7.(6) 8.(5) 9.(8) 10.(9) 11.(8) 12.(11) THe UNIVERSITY SCIENCE BULLETIN Key to the Species of Ptilomera KEY TO MALES Rear margin of metacoxa without thorn-like projection. P. (Proptilomera) himalayensis Hungerford and Matsuda (Himalaya), p. 414 Rear margin of metacoxa with thorn-like projection. P(Ritlomend) spp 3 Scena «ene peed ee Ge 2 Males. Middle femur with a fringe of hairs on its outer third at IVES: Vc) near ae i nae enc ope wien Mere Gi RarmT A CuNecd a Gig. oao-6-0'0 0 0 3 Females. Middle femur without such a fringe................. 28 Paramere bifurcate. P. wernert Hungerford and Matsuda (Philippines), p. 511 Paramere not bilurcate:@ ic. .c:5 sc ess ch ee ee 4 Dorsolateral projections of pygofer, as seen from above, extending half their length beyond lateral wings of suranal plate. P. agrioides Schmidt (India), p. 431 Dorsolateral projections of pygofer, as seen from above, much shorter, never extending half their length beyond lateral wings of suranal plate, often even shorter than lateral wings of suranal Plate ie Pes Wise Ses ees oe lesb, eh tiate case ane eer 5 Apical margin of pygofer broad, nearly truncate or depressed at mid- dle. Lateral margins of py gofer parallel or nearly parallel sided be- yond. lateral projections2s: (We)... os es eee 6 Apex of pygofer pointed, narrowly rounded or blunt. Lateral mar- gins of pygofer beyond its dorsolateral projections more or less strongly. comvergent. ...0.......) 240 8 Lateral wings of suranal plate well developed, reaching or extending beyond median lobe of suranal plate. ..P. chinai n. sp. (Borneo), p. 438 Lateral wings of suranal plate never reaching apex of median lobe caudally 2 ow Aaa es ae ery let eee en Ae ee a Median lobe of suranal plate broad. Seen in side view venter of py- gofer medially transversely constricted. Tip of pygofer not notched. Arm of paramere broadly spatulate at tip. P. cingalensis Stal. (Ceylon), p. 427 Median lobe of suranal plate not broad. Seen in side view, venter of pygofer not medially transversly constricted. Tip of pygofer notched. Arm of paramere not spatulate at tip. P. timorensis n. sp. (Timor Isl.), p. 436 Median lobe of sur anal plate large, clearly surpassing lateral wings Cautdaillty sc. ses os ek ve Se RE reas ah ne 9 Median lobe of suranal plate not large, slightly surpassing or not surpassing lateraltwines caudally... 7402 ..500100 eee 11 Paramere of usual shape as shown on pls. II] andIV............ 10 Paramere long and slender as shown on plate V. P. sarawakensis n. sp. (Borneo), p. 424 Arm of paramere as seen from the rear as shown on pl. ITI, figs. 3, 6. P. laticaudata (Hardwicke) (Nepal), p. 416 Arm of paramere as seen from the rear. pl. IV. fig. 6. P. assamensis n. sp. (Assam), p. 421 As seen from below pygofer conspicuously broad beyond its mid- GIGS tos he Sheahan ascn eas ie oa os aie Gee a eng rar a e al g 12 As seen from below pygofer not conspicuously broad beyond its mid- CG ie raisers. spss Mie ern Oa kc tucle arch Sina. AcackeNaet etn NN ae rr 14 Caudal half of that section of pygofer beyond its lateral projections parallel sided as seen from below. As seen laterally this distal por- tion is turned dorsally to a truncate tip as shown on pl. X, figs. 6, Test ok P. nunikanensis n. sp. (Indonesian Borneo), p. 441 13. (12) 14.(11) 15.(14) 16.(15) 17.(16) 18. (16) 19. (15) 20.(19) 23. (21) Genus PTILOMERA AMYOT AND SERVILLE Pygofer unusally broad beyond its dorsolateral projections, with sides converging to a dorsally directed point as shown on pl. XI, fig. (GY ay 0YG Fs oa Da BIE bFeaas abe peer atten abe ets Pele ere enna COMES cic lOO, A) als ge Raa 13 Lateral wings of suranal plate longer than broad, plainly surpassing median lobe, their rear margins longer than width of their tips. The paramere shape as shown on pl. XI, figs. 4, 7. Be kirkaldyi n. sp. (Borneo), p Lateral wings of suranal plate short, broad, only slightly surpassing median lobe, their rear margin not longer’ than the width of their tips. The paramere shape as shown on pl. XII, figs. 2, 3. IP. gressitti n. sp. (Borneo), p. Venter of pygofer with basal two-thirds narrow, resembling a broad keel. Sides above the keel concave to a position beyond the lateral projections of the pygofer. Venter of first genital segment with a well developed keel. (PI. XIII, fig. 3.) 407 p. 444 446 P. hylactor Breddin (Annam), p. 448 Venter of genital segments not as above...................... 15 Lateral wings of suranal plate with front and rear margins nearly qo eieallll eae teres ea ae ae iy oe Oe a ENS ee UD Pe en De aur re ep ea 16 Lateral wings of suranal plate with front and rear margins not par- GE Ce) bce ceca aes i ai ect ai CN ata Sine SR Ma ne re Rg 19 Lateral wing of suranal plate long, surpassing median lobe...... . 17 Lateral wing of suranal plate not surpassing median lobe....... 18 Median lobe of suranal plate short, but broader than width of lat- eral wing. Arm of paramere as seen in rear view with a ventral bulge or keel. (Pl. XIV, fig. 4.) P. breddini n. sp. (New Guinea), Median lobe of suranal plate longer than in above, but not wider than width of lateral wing. Arm of paramere as seen in rear view without the ventral bulge or keel. (Pl. XV, fig. 4.) P. papuensus n. sp. (Papua), p. Lateral wing of suranal plate short, its rear margin a little longer than the width of its tip. (PI. OQ fig. 2.) P. oribasus Breddin (Celebes), p. Lateral wing of suranal plate longer, its rear margin nearly twice as long as the width of its tip. (PI. XVI, fig. 7.) P. aéllo cheesmanae* n. subsp. (New Guinea), p. Median lobe of suranal plate plainly surpassing lateral wings. .. . 20 Median lobe of suranal plate not plainly surpassing lateral wings, 21 Exposed shaft of paramere broad and short. (Pl. X XVII, fig. 5.) P. harpalos Schmidt (Sumatra), p Exposed shaft of paramere more slender. (Pl. XXIV, fig. 5.) P. laelaps Breddin (Celebes), p. Lateral wings of suranal plate plainly surpassing median lobe... . 22 Lateral wings of suranal plate not plainly surpassing median lobe, 23 Exposed shaft of paramere longer than its arm. Hind coxa reach- ing rear margin of fifth abdominal tergite. (Pl. XX, fig. 3.) P. sumbaensis n. sp. (Sumba Isl.), p. Exposed shaft of paramere not as long as its arm. Hind coxa only reaching rear margin of fourth abdominal tergite. (Pl. XXI, fig. 3.) P. maai n. sp. (Borneo), p Lateral projection of pygofer indorsal view extending one-fifth its length beyond the lateral wing of suranal plate. Paramere long, slender, its shaft and arm in a continuous curve. (Pl. XXII, fig. 3.) P. pamphagus Breddin (Celebes), p. Lateral projection of pygofer shorter. Paramere shaft and arm not Sle ery er Ba NOM iis IEA Hae na cemihi arn h csiea ull Aveda ba eng Nacsige aL 24 p. 451 454 497 459 p. 493 488 478 * Male of P. aéllo aéllo Breddin is unknown to us. 408 24. (23) 25. (24) 26. (24) bo MN 29. (28) 30. (29) 31.(30) 32.(31) 33. (32) THE UNIVERSITY SCIENCE BULLETIN Distal half of that portion of pygofer beyond its lateral projections unusually slender, its caudal tip pointed.....................0. 25 Distal half of that portion of pygofer beyond lateral projections of usual shape and its caudal tip blunt..:..................:.... 26 Ground color of body and legs dark chocolate brown, nearly black, so dark that the black markings common to all species are all but obscured, even the longitudinal black bands of the front femur. Dis- tal half of dorsum of first genital segment abnormally elevated. (PI. XXIII, fig. 4.) Rear view of the arm of the paramere shows it straight, nearly parallel sided, and its tip rounded. (Pl. XXIII, fig. 3.) In dorsal view arm of paramere obscured by many long hairs. P. sumizome Esaki (8. O. Celebes), p. Ground color of body and legs reddish orange with the characteristic black markings of the genus. Distal half of dorsum of first genital segment not abnormally elevated. Rear view of arm of paramere shows it curved obliquely dorsad. Its sides not parallel. (Pl. XII, FESO Ox) shacscieeeean emis eater eek P. harpyia Schmidt* (Cambodia), p. Lateral wings of suranal plate slightly surpassing median lobe. P. dromas Breddin (syn. P. argus Breddin) (syn. P. asbolus Breddin) (Java), p. Lateral wings of suranal plate not surpassing median lobe....... 27 Arm of paramere as seen from the rear with a ventral keel as shown on pl. XXIV, figs. 4, 5. Exposed shaft of paramere longer than the CURVEG VAT IA fee AUN ile Nom se tes eth ate P. sumatranus n. s. (Sumatra), p. Arm of paramere as seen from the rear without a ventral keel. Ex- posed shaft of paramere not longer than the straight arm. P. lundbladi n. sp. (Sumatra), p. KEY TO FEMALES Rear margin of metacoxa without thorn-like projection. Without connexival spines. P. (Proptilomera) himalayensis Hungerford and Matsuda (Hima- laya), p. Rear margin of metacoxa with thorn-like projection. With connexi- Ways pIDeSh ese Gere eis ewe epee neni ne eee P. (Ptilomera) species 29 With a dorsally directed protuberance on the seventh abdominal ter- gite.........P. wernert Hungerford and Matsuda (Philippines), p. Without a dorsally directed protuberance on the seventh abdominal Gereiter sc. eae sae PET ee Bee ca ytd ate SNE Sn a 30 Without or with only the faintest indication of a ventrolateral lobe of- seventh abdominalisegment).);.2 0...) 2 2.3 ee ee 31 With a ventrolateral lobe of the seventh abdominal segment, which in one species might be overlooked as a basal notch in the dorso- lateral lober- (Pla XU figs 5:2) ceo ie Si ates a ee ee 38 Without conspicuous connexival spines. P. timorensis n. sp. (Timor Isl.), p. With conspicuous connexival spines...............++--+--+++: 32 Base of connexival spine arising from beneath the dorsal edge of the dorsolateral lobe of the seventh segment at its base and directed medially, then caudally.2 3. 2 okie a oa 33 Base of connexival spine not arising as above but beyond the seventh BOR IGS i ee wee See I UN CU) on aa RN OAs AY i eI ge 36 Dorsolateral lobe of seventh abdominal segment short, its tip only slightly surpassing the ninth tergite when visible. 481 466 503 485 488 414 511 436 P. agriodes Schmidt (syn. P. lachne Schmidt @ ) (8. India), p. 431 * A small male and female from Ceram Island, one of the Molluccas, ran out here but appear to be a subspecies developed on this island. See Pl. XIX............. P. harpyia ceramensis n. subsp. (Ceram Isl.), p. 470. 34. (33) 35. (34) 36. (32) 37. (36) 38. (30) 39. 40. (39) 41.(40) 42.(41) 43. (42) 44. (43) 45. (42) Genus PrmomenkA AMYOT AND SERVILLE Dorsolateral lobe of seventh abdominal segment long, its tip plainly surpassing ninth abdominal tergite when visible............... 34 Upper margin of dorsolateral lobe of seventh abdominal segment definitely curved downward. In side view appearing short. P. cingalensis Stal (syn. P. canace Schmidt) (Ceylon), p. Upper margin of dorsolateral lobe of seventh abdominal segment nearly straight. In side view appearing long..........:....... 35 In side view connexival spines are long, raised well above the dorsal margin of dorsolateral lobe of seventh abdominal segment. P. laticaudata (Hardwicke) (Nepal), p. In side view connexival spines usually shorter, flattened against or flattened slightly beneath dorsal margin of dorsolateral lobe of sev- enth abdominal segment.......... P. assamensis n. sp. (Assam), p. Sides of the vertically directed distal lobe of the seventh abdominal sternite not straight and parallel but curved and distally converging toward the truncate distal end. P. sarawakensis n. sp. (Brit. N. Borneo), Sides of the usually vertically directed distal lobe of the seventh ab- dominal sternite straight and nearly parallel. (If hidden by lateral lobesirumstinere) aoe Saipan ies dee teen Goran nance Ee gua Gussiieegelie 37 Dorsolateral lobe of seventh abdominal segment round and spine- like for its entire length as seen laterally. (Pl. XI, fig. 5.) P. kirkaldyi n. sp. (Borneo), Dorsolateral lobe of seventh abdominal segment starts as a triangu- lar lobe to become round, slightly sigmoid and spine-like. (PI. IX, LIDS Os) er ceee ni AARNE acannon ety P. china n. sp. (Borneo), Distal lobe of seventh abdominal sternite unusually large and not turned dorsally 9.2858 97 oy P. gressittt n. sp. (Borneo), p. Distal lobe of seventh abdominal sternite smaller and usually em- braced-by the lateral lobese ie ee eo Seventh abdominal segment with obviously short, often broad dorso- lateralelobess (Rl XOGV how. )e Wier ore rn ani oer aetees 40 Seventh abdominal segment with rather long, more or less spine-like dorsolateral lolbest 3 (RIRPXOXUING)) oe Ge i en 49 A narrow incission separating the dorsolateral and ventrolateral lobes of seventh abdominal segment.......................-.. 41 A wide incission, often shallow, separating the dorsolateral and ven- trolateral lobes of seventh abdominal seoment—) oss a shee a. 47 Incission separating the dorsolateral and ventrolateral lobes of sev- enth abdominal segment very shallow. (See Lundblad’s B in figure 136, pl. XXV.)..... P. lundbladi n. sp. (Subang Ajam, Sumatra), Imeissionedee per’ Ais cane oe Wee enn Ge ret ge eg te ne 42 Dorsolateral lobe definitely turned ventrad.................... 43 Dorsolateral lobe turned obliquely downward.................. 45 Color dark chocolate, almost black. Genital segments as shown on pl. XXIII, fig. 6..... P. sumizome Esqki (Southeastern Celebes), p. Colomtypicaltiorm theyzenusten tener y eae tae eee re 44 Dorsolateral lobe of seventh abdominal segment very broad at its base as shown on pl. XX, fig. 7. P. sumbaensis n. sp. (Sumba Isl.), Dorsolateral lobe of seventh abdominal segment narrow at base as shown on pl. XXII, fig. 4...... P. pamphagus Breddin (Celebes), p. In dorsal view dorsal margin of base of seventh abdominal connexi- vum nearly straight as shown on pl. XXIV, fig. 1. P. sumatranus n. sp. (Sumatra), p. In dorsal view dorsal margin of base of seventh abdominal connexi- vum short, curved outward, then inward to the base of connexival JONI BIST SOON OL) jollo OVI, 1A, I Boo oe eaoessuoodosdeusne do 46 409 42 ~I 416 421 p. 424 p. 444 p. 488 446 p. 488 481 p. 474 478 485 410 46. (45) 47.(40) 48. (47) 49. (39) 50. (49) 51.(50) 52. (50) 54. (53) 55. (54) THe UNIverRSITY SCIENCE BULLETIN Dorsal margin of dorsolateral lobe of seventh abdominal segment nearly straight or slightly concave and shorter than connexiva! spine. (Pl. XVI, fig. 9.) P. aéllo aéllo Breddin (New Guinea), p. Dorsal margin of dorsolateral lobe of seventh abdominal segment convexly curved and longer than connexival spine. (Pl. XVI, fig. 10.) P. aéllo cheesmanae N. subsp. (Dutch New Guinea), p. Connexival spines stout, plainly shorter than seventh tergite, not half as long as upper margin of dorsolateral lobe from base of con- nexival spine to its tip. (Pl. X XVI, fig. 7.) P. laelaps Breddin (Celebes), p. Connexival spines longer, at least half as long as upper margin of dorsolateral lobe from base of connexival spine to its tip......... 48 Dorsolateral lobe slender as shown on pl.X XVIII, fig. 6. Connexi- val spine nearly as long as upper margin of dorsolateral lobe from base of connexival spine to its tip... P. oribasus Breddin (Celebes), p. Dorsolateral lobe stout as shown on pl. X XVII, fig. 7. Connexival spine half as long as upper margin of dorsolateral lobe from base of connexival spine to its tip....... P. harpalos Schmidt (Sumatra), p. Dorsolateral lobe of seventh abdominal segment long, stout and straight, its upper margin from base of connexival spine to its tip nearly twice as long as connexival margin of seventh abdominal seg- ment. (Pl. XIII, fig.6.).......... P. hylactor Breddin (Annam), Dorsolateral lobe of seventh abdominal segment only moderately long, its upper margin from base of connexival spine to its tip much less than twice as long as connexival margin of seventh abdominal SCCMME ML Me een Ae MS bey aetna Set ae sh Re 50 Incission between the dorsolateral and ventrolateral lobes of seventh abdominal segment at least as deep as broad.................. 51 Incission between the dorsolateral and ventrolateral lobes of seventh abdominal segment not as deep as broad...................... 52 Incission between the dorsolateral and ventrolateral lobes of seventh abdominal segment definitely deeper than broad, upper lobe stout. GRIEVE io 05) etnias Pere eee P. papuensis n. sp. (Papua), p. Incission between the dorsolateral and ventrolateral lobes of seventh abdominal segment rounded. (PI. XIV, fig. 7.) Upper lobe rather slenderres eee Ge een area: P. breddini n. sp. (New Guinea), p. Ventrolateral lobe of seventh abdominal segment faintly indicated as Showmvon oll DOxG sion Gre ens ee eee P. maai n. sp. (Borneo), p Ventrolateral lobe of seventh abdominal segment plainly indica- L210 UAE ete BN ot ag aaa eae COR res ea eu MarR aEinereGtnia Gidha. o-o\o.0 ¢ 53 2) Ventrolateral lobe of seventh abdominal segment narrow spine-like. (RIREXG fies») ree P. nunikanensis n. sp. (Indonesian Borneo), p Vanerolareral lobe of seventh abdominal segment triangular, blunt tipped orttruncate ss on er Oe Ge ea ele eee 54 Incission between the dorsolateral and ventrolateral lobes very broad and shallow Ventrolateral lobe triangular. P. dromas Breddin (syn. P. argus Breddin) (syn. P. asbolus Breddin) (Java), Incission between the dorsolateral and ventrolateral lobes not so broad and shallow. Ventrolateral lobe blunt or truncate........ 55 In side view the ventrolateral lobe of seventh abdominal segment shorter than the venter of the seventh abdominal segment before it. In ventral view the ventrolateral lobe is not broad. (Pl. XXVIII, figs. 8,9.) (The female type)...... P. dorceus Breddin (Celebes), p. In side view the ventrolateral lobe of seventh abdominal segment longer than the venter of seventh abdominal segment before it. In ventral view the ventrolateral lobes of types of both P. tagrina and P. harpyia are truncate at tips. However, the latter species seems tO Vary asishowm on) pl xeViil ies sa) tOlOr in. case h-eelde) eee cn 56 458 459 488 497 493 p. 448 454 451 p. 476 p. 441 p. 503 501 Genus PTILOMERA AMYOT AND SERVILLE All 56.(55) The dorsolateral lobe rather narrow at base as shown on pl. XVII, figs. 2,4. Ventrolateral lobes broad. As seen from below, their inner margins almost touching or overlapping and their anterior margin turned mesally.......... P. tigrina Uhler (Hong Kong, China), p. 462 The dorsolateral lobe broad at base as shown on pl. XVIII, figs. 2, 5, 7, 8, 9. Ventrolateral lobes not as above. P. harpyia Schmidt* (Cambodia), p. 466 Besides those we have keyed out above we have a female from “Oek Simakock, N. Habinsoram, Sumatra July 3, 1931 V. D. Meer Mohr” that looks like Lundblad’s E in his fig. 136, and a male from the same place but on Aug. 4, 1931. They are none of the species we have keyed out. Lundblad’s specimen came from Belige, W. Sumatra. Since males are quite constant in their combination of characters and can be satisfactorily keyed out we believe that with careful study there will be recognized more species rather than fewer species in this difficult genus. In some species the females vary so slightly that they are readily keyed out but there are others that appear to be so variable that they make trouble when we try to place them in a key. We have done our best to enable one to identify the females encountered, but we are conscious of its shortcomings and have pro- vided illustrations of all species. Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda Plate II 1958. Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda, Bull. Brooklyn Ent. Soc. 53(5):117-128, 2 Pls. (Described new subgenus and species from Singla Darjiling. E. Himalayas. ) 1960. Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda, Univ. Kansas Sci. Bull. 41(1):7, 17 fig. 33. (subgenus Proptilomera. ) 1960. Ptilomera (Proptilomera) himalayensis, Matsuda, Univ. of Kansas Sci- ence Bull. 41(2):267, 537 figures 539, 609, 614, 615, 620, 624. Redescription of subgenus Proptilomera More primitive group of the genus Ptilomera, represented by P. himalayensis. It lacks the metacoxal thorn-like projection that is present in all other species. The seventh abdominal and genital segments of both sexes are simple and more primitive than in other species of the genus. The seventh abdominal tergite short in both sexes. First genital of male short, broad without a ventral keel. The pygofer also short and broad, lacking a distinct dorsolateral projection. Paramere simple, not divided into a shaft and arm. * A small male and female from Ceram Island, one of the Molluccas, ran out here but appear to be a subspecies evolved on this island. See Pl. XIX. P. harpyia ceramensis n. subsp. (Ceram), p. 470. Whether P. tigrina and P. harpyia are one variable species or distinct species must await the discovery of males of P. tigrina accompanied by females. 412, THe UNIVERSITY SCIENCE BULLETIN PLATE II Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda Fic. 1. Apterous female from Kurseong, Himalayas (Brit. Mus.). Note the absence of metacoxal thorn-like spine and connexival spine. Fic. 2. Apterous male from Kurseong, Himalayas (British Mus.). Note both the absence of metacoxal thorn-like spine and the very large first genital segment that almost covers the second genital segment. Fic. 3. Suranal plate of the male paratype, showing the elongate median lobe and the curious hook-like wings of suranal plate. Fic. 4. Macropterous male holotype. Fic. 5. Macropterous female allotype. Dorsal view of caudal abdominal segments. Note the large seventh abdominal tergite and the ends of the median lobe of the seventh ventrite on either side of the anal segment. Fic. 6. Macropterous female allotype. Side view of caudal abdominal seg- ments to show the curious shape of the median caudal lobe of the seventh ab- dominal ventrite. Fic. 7. Male holotype. Side view of caudal abdominal segments showing the curious lateral wing of suranal plate which is nearly covered by large first genital segment, the short pygofer and the erect paramere. Fic. 8. Female allotype. Ventral view of the caudal abdominal segments. Note the median longitudinal keel on the distal lobe of the seventh abdominal ventrite and the caudolateral projections, the tips of which are shown in figure 5 above. Fic. 9. Holotype male, ventral view of genital segments. Fic. 10. Paratype male, side view of pygofer, suranal plate, paramere and endosoma. Genus PrmoMERA AMYOT AND SERVILLE PLATE II 413 414 THe UNIVERSITY SCIENCE BULLETIN Suranal plate not typically three lobed and entirely exposed, but hidden beneath dorsum of first genital, median lobe comprising most of the suranal plate, the lateral wings slender, not lobe-like, with only their curved tips exposed beyond the caudolateral margin of the first genital segment. Female without connexival spines. Seventh segment without dis- tinct dorsolateral or ventrolateral lobes on its rear margin which is straight. Median caudal lobe of seventh sternite with a distinct median longitudinal ridge, its lateral margins converging then turned obliquely outward to produce on either side a flat caudo- lateral projection, the ends of which are visible in dorsal view. Description of Ptilomera (Proptilomera) himalayensis Size: Male holotype (macropterous): Length including wings 14.4mm.; width of head 1.64mm.; width of anterior lobe of pro- notum 2.1 mm.; width across humeri 2.8mm.; greatest width of body across mesoacetabula 3.48 mm. Female allotype (macropterous): Length including wings 14.44 mm.; width of head 1.68 mm.; width of anterior lobe of pronotum 1.93 mm.; width across Tannen 2.81 mm.; “preatest width of body across cnenoeraiclnnle: 3.90 mm. Male (apterous): Length 10.48 mm.; width of head 1.7 mm.; width of pronotum 2.00 mm.; length of PyeconGtunn 2.80 mm.; great- est width of body across negonveottclonle 3.48 mm. Female (apterous): Length 12.48 mm.; width of head 1.83 mm.; width of pronotum 2.05 mm.; length of meconoann 2.95 mm.; aoa est width of body across mecosectabela 405mm. We have) seen a macropterous male that is 15.5 mm. long including wings, and an apterous male only 10.1 mm. long. Color: Apterous forms have the color and pattern typical for the genus except that the black band on the side of the mesothorax is narrower, undulate, and not divided. Mesothorax with a lateral black band, broader in front, and reaching back onto metacetabula; mesoacetabula with a black spot. Venter pale testaceous, covered with short silvery pile. Abdominal tergites of deilated male with black bands covered with silvery pile across bases of first four seg- ments. Antennae brown. Legs testacenes to reddish brown; an- terior femora with two longitudinal reddish brown bands; middle and hind femora pale testaceous at tips. The macropterous forms: reddish brown to testaceous above, wings brown. Head with two or more indefinite embrowned spots. Prothorax with a black spot on anterior median margin, a transverse Genus PrmOMERA AMYOT AND SERVILLE A415 band on anterior margin of posterior lobe broken in the middle by a low testaceous carina; sides of anterior lobe with a longitudinal dark band not reaching to eyes and a large spot in proacetabula. Structural Characteristics: Relative length of antennal segments of male holotype. Ist:2nd:3rd:4th::125:30:44:25, of the female al- lotype, Ist:2nd:3rd:4th::115:27:40:?. The last segment of the holo- type somewhat curved and flattened. Actual Lengths of Leg Segments Ist tarsal 2nd _ tarsal Front leg Femur Tibia segment segment Male (holotype) ........ 7.43 mm. 6.19 mm. 2.85 mm. 1.42 mm. Male (apterous) ........ 6.81 mm. 5.71 mm. 2.76 mm. 1.29 mm. Female (allotype) ...... 6.29 mm. 5.24 mm. 2.38 mm. 1.24 mm. Female (apterous) ...... 6.67 mm. 5.71 mm. 2.52 mm. 1.29 mm. Middle leg Femur Hind leg Femur Male (holotype) ........ 19.50 mm. Male (holotype) ...... 25.37 mm. Male (apterous) ........ 19.05 mm. Male (apterous) ...... 22.60 mm. Female (allotype) ....... 16.90 mm. Female (allotype) ..... 19.60 mm. Female (apterous) ...... 17.14 mm. Female (apterous) .... 20.48 mm. In the hind leg of the deilated Male paratype which is larger than the holotype, the femur is 28.09 mm. long. The front femur of the male is rather stout and has a characteristic shape. The front femur of female is normal in shape and provided with a small black protuberence ventrally near distal end. This protuberence is larger in the male. Tibia of front leg also has three small protuberences on its inner base; these protuberences are smaller in female than in male. Male middle femur with densely ciliated ventral margin confined to distal two-fifths; basal three-fifths bare; metacoxal spine represented by a blunt black protuberance. Length of pregenital segments: genital segments::60:30. Seventh abdominal tergite of male short, not longer than the fifth and sixth together. Male genital segments: first genital segment broad and short; without a ventral keel. Suranal plate small. In normal position almost en- tirely hidden, its median lobe obscured by the dorsum of first genital, its lateral lobes slender and hidden except for their tips. (See Plate II fig. 3.) Pygofer simple, short and broad, without dorsolateral projections of all other species. Paramere as shown on Plate II fig. 10. Female. Length of thorax:abdomen::102:99. Seventh abdomi- nal segment is also simple and quite primitive. Seventh abdominal tergite short, only slightly longer than sixth, without connexival spines. In dorsal view the black projection one on each side near the caudal apex of the insect are the tips of prolongations of the seventh sternite, which in other species of Ptilomera is a single flap 416 Tue University SCIENCE BULLETIN that is nearly truncate across its end. In side view there are only slight indications of lateral lobes so striking in other species of this genus. Types: The macropterous male holotype, macropterous female allotype and a deiilated male paratype bear the label “Singla Dar- jiling, E. Himalayas June 13. 1500 ft. Ld. Carmechael” in the Francis Huntington Snow Coll., The University of Kansas. Distribution: In addition to the types we have seen the follow- ing: “Mungphu” 2 macropterous males (British Mus.); “Ind Mus. Kurseong Alt. 3200 ft. E. Himalayas. 2, 6. IHII-10. F. Gravely.” apterous male and female. (British Mus.); “N. India” 1 dealated male. (British Mus.) “Sikkim, Gopaldhara, Rungbong Vall. H. Stevens.” two apterous male, one apterous female. (British Mus. ). These British Museum specimens bear a manuscript name, “Ptilo- mera primitiva Lundblad” and were seen there by the junior author (1960) who recognized them as our Ptilomera (Proptilomera) himalayensis Hungerford and Matsuda and were borrowed so that we could describe the apterous forms. It is evident that Dr. Lund- blad before 1933 recognized that this is a primitive species and his name would have been more meaningful than our own. Ptilomera (Ptilomera) laticaudata (Hardwicke ) Plate III 1825. Gerris laticaudata Hardwicke, Trans. Linn. Soc. 14:134 Tab. VI figs. 1-4. (Described from “Nepaul’, His figure 2 is an adult male not a nymph and figure 4 incorrectly drawn. Types in British Museum. ) 1843. Ptilomera laticauda (Hardwicke), Amyot et Serville Hem. 413-414. P. VIII fig. 3. (Described the genus but misspelled laticaudata and added Java in error.) — 1853. Ptilomera laticauda, Herrich-Schiffer, Wans. Ins. 9:65 fig. 940. (Un- known to him but copied from Amyot and Serville). 1866. Ptilomera laticauda, Mayr, Novara-Exped. Zool. Theil Bd. 2(1):177. (Followed Amyot and Serville in making P. cingalensis Stal a synonym and questioned P. tigrina Uhler). 1900. Ptilomera laticaudata, Horvath, Semon Zoologische Forschungsreisen in Australien und dem Malayischen Archipelagos, Denkschriften VIII 639, (Reports from Buitenzorg, Java. Surely not above species). 1904. Ptilomera laticaudata, Distant, Fau. Brit. Ind. Rhynch. 2:185. (Only first three references above apply to this species. He was in error in making synonyms of P. cingalensis Stal and P. tigrina Uhler. His figure 133 is not of P. laticaudata (Hardw.) and the drawing he labeled fe- male is a male which looks like P. agriodes Schmidt). 1927. Ptilomera laticaudata, Esaki, Eos, Rev. Esp. Ent. 3(3):258 (questions all distributional records except Nepal). 1933. Ptilomera laticaudata, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 373, 418, 422-423. fig. 133. (Saw and redescribed the type and gave drawings of male genitalia). Genus PrmoMERA AMYOT AND SERVILLE 4I7 1958. Ptilomera laticaudata, Hungerford and Matsuda, Bull. Brooklyn Ent. Soc. 53(5):117-118, Pl. I, Pl. Il G-H. (Described and figured female and thought this to be this species. ) 1960. Ptilomera laticaudata, Matsuda, Univ. of Kansas Sci. Bull. 41(2):267, 269, 587, 541, (Fig. 619 on p. 537 not laticaudata). The following references are misdeterminations: 1892. Ptilomera laticaudata, Kirby, Linnean Soc. Jour. Zool. 24:124. (Not P. laticaudata (Hardw.) but P. cingalensis Stal.) 1899. Ptilomera laticaudata, Kirkaldy, Ann. Soc. Ent. Belg. 43:508. (Re- corded a female from “S. Celebes, Samanga. (Fruhstorfer Nov. ’95.)” In Kirkaldy Collection we find 3¢ ¢ and 29 2 from this place. They are P. laelaps Breddin). 1901. Ptilomera laticaudata, Breddin, Abh. Naturf. Ges. Halle Bd. 24:20. (Quoted Kirkaldy as 1899. ) 1918. Ptilomera laticaudata, Paiva, Records of Indian Museum 14:24 (Prob- ably P. agriodes Schmidt. ) 1919. polos laticaudata, Paiva, Records of Indian Museum 16, Pt V, No. 23; 365. For more than one hundred years this species from Nepal re- mained indistinguishable in the literature. Then in 1933 Dr. Lund- blad saw the types, an apterous male and a macropterous male with broken wings, in the British Museum. He redescribed the types and gave drawings of the male genitalia. In 1958 we described a lone female from Darjiling, just east of Nepal that we thought might be the female of P. laticaudata (Hardw.) but stated that the determination could not be certain until both sexes were taken to- gether from one place. Now we have found, in the Collection of the British Museum, one male and two females from Sikkim. The females are like our neallotype of P. laticaudata from Singla, Dar- jiling, East Himalayas and the male is like the two male types of P. laticaudata in the British Museum. Therefore, we now know both sexes of P. laticaudata (Hardwicke). A closely related species from Assam is described after this species. Size: Although Hardwicke gave the length of this species as eleven lines, Lundblad (1933) said the type male is relatively small and around 14mm. long. Since the genitalia in both male types have now been removed and the identification confirmed, we ac- cept the length given by Lundblad. Mr. Izzard of the British Museum has obligingly sent us the following measurements: Width across the head including eyes 2.73mm.; width of pronotum 2.73 mm.; length of mesonotum (apterous form) 3.73mm. Our neallotype was also described without some of the measurements we are giving in this paper. Here we give them for the Female. Length 14.19mm.; width of head 1.95mm.; width of pronotum 2.24 mm.; length of mesonotum 3.19mm.; greatest width of body across mesoacetabula 3.95 mm. 418 THe UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) laticaudata (Hardwicke ) Fic. 1. Male from Sikkim. Compared with type of Ptilomera laticaudata. _ Fic. 2. Male. Ventral view of genital segments of the above specimen from Sikkim. Fic. 3. Rear view of right arm of paramere of P. (Ptilomera) laticaudata (Hardwicke ) from Sikkim. Fic. 4. Female neallotype of Ptilomera laticaudata by Hungerford and Matsuda. Fic. 5. Female neallotype, dorsal view of sixth and seventh abdominal seg- ments. Fic. 6. Male. Paramere of a type drawn from a slide mount by Matsuda at the British Museum. Fic. 7. Male type. Suranal plate drawn from a slide mount by Matsuda. Fic. 8. Female neallotype. Side view of sixth and seventh abdominal seg- ments. Genus PrmoMERA AMYOT AND SERVILLE PLATE III 419 420 Tue UNIVERSITY SCIENCE BULLETIN Color: Typical for the genus. However, all three females known to us have four black dots on the vertex. Structural characteristics: The male types now lack antennae. However, if Hardwicke’s fig. 3 is correctly drawn then the relative length of antennal segments of the Male are Ist:2nd:3rd:4th::34:10: 12:7+. Those of our neallotype 125:25:37:?. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Malesty pers cn eee. eae 7.33 mm. 5.20 mm. 3.66 mm. 1.66 mm. Female neallotype (apterous): (2. ees: 6.57 mm, 5.62 mm. 3.00 mm. Female Sikkim, apterous)! cece 6.76 mm. 5.81 mm. 3.43 mm, 1.52 mm. Middle leg Femur Hind leg Femur Malestypezs Sener a nasar 25.00 mm. Malestype * = = eee 30.00 mm. apterous) a5 joe oe 17.14 mm. Female neallotype Female Sikkim, (apterous) see ae 20.00 mm. (apterous; se eure 18.19 mm. Female Sikkim, (apterous) ie eee 20.38 mm. Length of pregenital abdominal segments: Genital segments: :73: 57 (from Sikkim). Male genital segments: Suranal plate with median lobe very broad but short, not as long as the distal width of a lateral wing, yet surpassing the latter caudally (unlike the drawing in Hardwicke’s Fig. 4). Front and rear margins of lateral wings not parallel. Pygofer relatively short, its tip bluntly pointed, dorsolateral projections stout, short and bluntly rounded at tip. Paramere in dorsal view slightly sigmoid, the union of shaft and arm indistinct. In rear view the dista half of arm as shown in Lund- blad’s figure 133 and as fig. 3 of Plate III. Female neallotype. Length of thorax: Abdomen: :120:136. Seventh abdominal tergite with rear margin not appearing broadly rounded. Base of connexival spine arising from beneath the dorsal edge of seventh connexivum, the transverse base of connexival spines over- lapping the rear margin of seventh abdominal segment on either side. Connexival spines long, their tips attaining the tips of dorso- lateral lobe. Where the transverse base of the spine turns caudally the spine turns obliquely upward for about one-third of its length before turning caudally. Without or with only the faintest indica- tion of a ventrolateral lobe of the seventh abdominal segment. Dorsolateral lobe of seventh abdominal segment long, its upper mar- gin nearly straight. * Measurements of the male type were made by Mr. R. J. Izzard of the British Museum. Genus PrmomMERA AMYOT AND SERVILLE 49) A D C Fig. 133. Ptilomera laticaudata Hardw. 6. A Genitalkapsel von oben; B linker Genitalgriffel; C Griffelende von der Spitze aus gesehen; D Analplatte. Nach der Type aus Nepal. Fic. 1. Dr. Lundblad permitted us to use this figure which illustrates the male genital characters of one of the types in the British Museum. Comparative notes: We have long considered some specimens from Assam as this species. However, we now can separate both sexes of it from P. laticaudata (Hardw.) and regard the specimens from Assam as a new species. Types: Two male types from Nepal are in the British Museum. One neallotype from “Singla Darjiling, E. Himalayas, June 13, 1500 ft. Ld. Carmichael” is in the Torre-Bueno Collection at K. U. Distribution: In addition to the above we have seen two females and one male bearing the label “Sikkim, Gopaldhara, Rungbon. Vall. H. Stevens.” These are like the neallotype above and are in the British Museum. Ptilomera (Ptilomera) assamensis n. sp. Plate IV Size: Male, (apterous): Length 13.8 mm.; width of head 2.00 mm.; width of pronotum 2.33 mm.; length of mesonotum 3.29 mm.; great- est width of body across mesoacetabula 3.81 mm. Female (apterous): Length 14.29 mm.; width of head 2.05 mm.; width of pronotum 2.19 mm.; length of mesonotum 3.19 mm.; great- est width of body across mesoacetabula 3.90mm. As we have stated in “Taxonomic Problems Involved,” we will not hereafter give figures or size variation, but the following will illustrate the problem by this species. Length of apterous males ranges from 13 to 15 mm.; width of head 1.85 to 2.00 mm.; width of prothorax 1.97-2.27 mm.; THE UNIVERSITY SCIENCE BULLETIN PLATE IV Fic. Fic. Fic. Fic. Fic. Fic. Fic. ND UR Ob Ptilomera (Ptilomera) assamensis n. sp. Female, dorsal view. Female, dorsal view of sixth and seventh abdominal segments. Male, ventral view of genital segments. Male, dorsal view. Female, ventral view of sixth and seventh abdominal segments. Arm of right paramere as seen from the rear. Female, dorsal view of sixth and seventh abdominal segments. This specimen ‘has a hand written label “Karwapam.” Its connexival spines are unusually long. Fic. 8. Female, an oblique view to show the flat lying connexival spines. Fic. 9. Female. side view of sixth and seventh abdominal segments. Genus PritoOMERA AMYOT AND SERVILLE 493 greatest width of body across mesoacetabula 2.73-3.81 mm. Length of apterous females ranges from 13.90-14.88 mm.; width of head 1.89 to 2.05 mm.; width of prothorax 1.87-2.14mm.; greatest width of body across mesoacetabula 3.36 to 3.99 mm. There are other species with even greater variations in size. Color: Ground color yellowish to reddish brown, abdominal tergites, except the last one, may be black in the males or brown with only the lateral margins and intersegmental lines black, con- nexivum brown. Clypeus black, two narrow black converging lines on vertex. Anterior median margin of pronotum with two black spots or a continuous connecting line between them behind the interocular space. Anterior margin of mesonotum with a lateral black spot on either side. Sides of mesothorax usually with two longitudinal undulate dark bands separated by a band of silvery pile above a yellow streak. The upper band black, the lower one dark brown and sometimes incomplete. Venter pale yellow, cov- ered by a short silvery pile. Antennae, front tibia and tarsus dark brown. Front femur with three dark brown to black longitudinal bands; legs elsewhere light brown. Structural characteristics: Relative length of antennal segments. Most specimens have lost at least the last two segments. There is a male in the British Museum labeled “Assam, W. F. Badgley 1906- 185.” This has the left antenna entire and measures: Ist:2nd:3rd: Ath: :175:37.5:55:32.5. and a female from the same place with the right antenna entire, Ist:2nd:3rd:4th: :132.5:28.5:43:28.5. Actual Lengths of Leg Segments 1st tarsal 2nd tarsal Front leg Femur Tibia segment segment IMIG = heir eng aes = ere on 8.33 mm. 7.14 mm. 3.86 mm. 1.81 mm. Female 32 ee 7.38 mm. 6.33 mm. 3.71 mm. 1.52 mm. Middle leg Femur Hind leg Femur Ni alee tinea ene sot oes 21.66 mm. Maleypebester epee 27.38 mm. Bemalessee seo ee ow 19.87 mm. Hemaleweenen en ee 22.97 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::77:48. Genital segments: Suranal plate with median lobe broad, but as long as or longer than the distal width of a lateral wing and more obviously surpassing the lateral wings caudally than in P. laticaudata (Hardw.); front and rear margins of lateral wings not parallel. Pygofer relatively short, its tip roundly pointed, its dorsolateral projections short and broad at base as seen ventrally. Paramere with much of its shaft hidden beneath the median lobe, its exposed part superficially resembling that of P. laticaudata. 494 THe UNIVERSITY SCIENCE BULLETIN However, the distal half of the arm as seen from the rear lacks the definite constriction that Lundblad showed for P. laticaudata (Hardw.). Female. Length of thorax:Abdomen::121:143. Seventh abdomi- nal tergite with rear margin appearing broadly rounded and much shorter than the sixth tergite. Base of connexival spine arising from beneath the base of seventh connexivum as a swelling that leaves no connexival edge. The obliquely turned base of connexival spine but slightly overlaps the caudolateral edge of the seventh tergite, then turns caudally and lies flat to its tip which seldom reaches tip of dorsolateral lobe. Without or with only the faintest indication of a ventrolateral lobe of the seventh abdominal segment. Dor- solateral lobe of seventh abdominal segment long but not a straight line. Comparative notes: A close relative of P. laticaudata but sepa- rated from it as shown in the key. Types: Male holotype and female allotype and one male and one female paratype from “Assam. W. F. Badgley 1906” and one female paratype “Karwapaui 1907” are in the British Museum. We have in the K. U. Collection the following paratypes: 43 3,39 9 from “Assam, India,” and3 g ¢ and1 ¢@ labeled “Assam” that came to us in the Kirkaldy Collection. Distribution: Known only from the type localities. Ptilomera (Ptilomera) sarawakensis n. sp. Plate V Size: Male holotype (apterous). Length 16.19mm.; width of head 2.38 mm.; width of pronotum 2.48 mm.; length of mesonotum 3.76 mm.; greatest width of body across mesoacetabula 4.19 mm. Female allotype (apterous). Length 17.71 mm.; width of head 2.29mm.; width of pronotum 2.38mm.; length of mesonotum 3.90 mm.; greatest width of body across mesoacetabula 4.45 mm. Male (macropterous). Length with wings 17.38 mm.; width of head 2.29mm.; width across humeri 3.33mm.; greatest width of body across mesoacetabula 3.86 mm. Female (macropterous). Length of body without wings 16.90 mm.; width of head 2.19 mm.; width across humeri 3.05 mm.; greatest width of body across mesoacetabula 3.86 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male holotype Ist:2nd:3rd: 4th: :180:40:59:37. Female allotype 1st:2nd:3rd:4th: :162:35:48:38. Genus PrrLOMERA AMYOT AND SERVILLE 425 PLATE V Ptilomera (Ptilomera) sarawakensis n. sp. Fic. 1. Female allotype, dorsal view. Fic. 2. Male holotype, ventral view of genital segments. Fic. 8. Male holotype, dorsal view. Note the long first genital segment, the median lobe of suranal plate and the long parameres. Fic. 4. Female allotype, ventral view of thee sixth and seventh abdominal segments. Fic. 5. Female allotype, side view of sixth and seventh abdominal seg- ments. 496 THe UNIVERSITY SCIENCE BULLETIN Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Malect ax 9) eae oe Foe ee 9.42 mm. §.24 mm. 4.76 mm. 1.86 mm. Remalewin cee aes wees 8.91 mm. 7.71 mm. 5.1 mm. Middle leg Femur Hind leg Femur Male sir Pre cau Pgs ae 26.1 mm. Male. ale) ae, oe eee 30.48 mm. Hemale, cat shorn owe hs 23.47 mm. Hemale! 22) — Se eee 27.86 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::87:67. Suranal plate with median lobe rather broad and extending caudally beyond the short lateral wings which have their front and rear margins not parallel. Pygofer of usual shape. Dor- solateral projections, as seen from below, broad at base but nearly pointed; as seen from above about as long as lateral lobe of suranal plate. Paramere long, shaft and arm not distinguishable. See plate 5, fig. 2. Female. Length of thorax:Abdomen::145:180. Seventh abdomi- nal tergite nearly as long as sixth tergite (18.5:17.8). Connexival spine moderately short, arising from end of the curved edge of the seventh connexivum, thus the connexival spines have a tendency to cross each other, without or with only the faintest indication of a ventrolateral lobe of seventh segment; dorsolateral lobe of seventh abdominal segment long. Sides of the vertically directed distal lobe of seventh abdominal sternite not straight and parallel, but curved and distally converging toward the truncated distal end. Comparative notes: This species runs out with P. laticaudata (Hardw.) in both keys to the sexes, couplet 9 for males and couplet 32 for females. Types: Holotype apterous male, allotype apterous female bear the label “Sarawak: Mt. Dulit R. Koyan 2500 ft. Primary Forest Nov. 13, 1932.” “On surface of fast water.” Also a macropterous male bearing the above label; a macropterous female from the above lo- cality but taken at 850 ft. on Aug. 22, 1932; an apterous female taken “skating on slow moving water at the junction of Tinjar and Lejok rivers at foot of Mt. Dulit Aug. 15, 1932; another apterous female “on fast parts of stream on Dulit Trail Aug. 16, 1932.” All of the above are labeled “paratypes” and were taken by B. M. Hobby and A. W. Moore on the Oxford Univ. Exp. The types are in the British Museum. Distribution: Known only from the above type localities. The finding of a specimen “on slow moving water” is only known such record. Ptilomera are usually recorded from fast moving water as were the other specimens of this type series. Genus PrmomMERA AMYOT AND SERVILLE 427 Ptilomera (Ptilomera) cingalensis Stal Plate VI 1855. Ptilomera cingalensis Stal, Ofvers af K. Vet. Akad. Forh. 12, 190 (De- scribed from “Ins. Taphrobana” an ancient name of Ceylon). 1927. Ptilomera cingalensis, Esaki, Eos Rev. Esp. Ent. 3(3):259. 1933. Ptilomera cingalensis, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 378, 421. 1933. Ptilomera canace Schmidt, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12: 372, 373, 421 (Says is a synonym of P. cingalensis Stal, with which we agree). 1960. Ptilomera cingalensis, Matsuda, Univ. Kansas Sci. Bull. 41(2):267. Since P. canace Schmidt 1926 is a synonym of P. cingalensis Stal 1855 the fol- lowing references belong here: 1926. Ptilomera canace Schmidt, Ent. Mitt. 15(1):65 (Described male and female from Ceylon). 1927. Ptilomera canace, Esaki, Eos Rev. Esp. Ent. 3(3):259. 1960. Ptilomera canace, Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 539 fig. 626. Dr. Lundblad found the type, a female in the Riksmuseet in Stock- holm and also a male that Stal did not describe. However, Lund- blad did not describe either sex but stated that it is a species that can be recognized as such. We have not seen the type of P. cingalensis but since Lundblad placed P. canace Schmidt as a synonym of P. cingalensis Stal (they both came from Ceylon) and we have studied four types of P. canace Schmidt (1 male and 1 female with their genital segments mounted on slides and two others). We know our specimens from Ceylon are P. cingalensis and offer the following description of the species. Size: This ceylonese species is indeed variable in size. The males ranging from 12.14mm. to 16.66 mm. in length, and females from 12.0mm. to 15.0mm. The pair we have chosen from our Kirkaldy Collection have the following measurements: Male (apterous): Length 16.66mm.; width of head 2.40 mm.; width of pronotum 2.67 mm.; length of mesonotum 3.43 mm.; great- est width of body across mesoacetabula 4.38 mm. Female (apterous): Length 14.0mm.; width of head 2.1 mm.; width of pronotum 2.1 mm.; length of mesonotum 2.9 mm.; greatest width of body across mesoacetabula 3.86 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments: Male from Pundaluoya, Ceylon, 1st:2nd:3rd4th: :160:40:50:34. Female from Pundaluoya, Ceylon, Ist:2nd:3rd:4th: :138:32:41:30. The male type of P. canace Schmidt has only two basal antennal segments: Ist:2nd::154:40. One of our males from Pundaluoya, 498 ‘The UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) cingalensis Stal Fic. 1. Female from Pundaluoya, Ceylon. Dorsal view. Fic. 2. Male from Pundaluoya, Ceylon. Ventral view of genital segments. Fic. 38. Female from Pundaluoya, Ceylon. Ventral view of sixth and — seventh abdominal segments. Fic. 4. Male from Pundaluoya, Ceylon. Dorsal view. Note carefully the shape of suranal plate, caudal end of pygofer and shape of paramere. Fic. 5. Right side view of genital segments of above male. Fic. 6. Ptilomera canace Schmidt, female type. abdominal segment drawn from the slide mount. Fic. 7. Ptilomera cingalensis Stal. Female from Pundaluoya, Ceylon. Left side view of sixth and seventh abdominal segments. Fic. 8. Ptilomera canace Schmidt. Male type. drawn from the slide mount. Dorsal view of seventh Last genital segment 429 Genus PrmoMERA AMYOT AND SERVILLE PLATE VI 1367 17 430 THe UNIVERSITY SCIENCE BULLETIN Ceylon with only two basal segments Ist:2nd::178:43 which shows a considerable variation. Actual Lengths of Leg Segments (Pair from Pundaluoya, Ceylon) Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment hall attr ee os i erate haat 9.43 mm. 8.19 mm. 4.95 mm. 1.81 mm. Hémalesie he 2 eee 7.00 mm. 6.24 mm. 3.57 mm. 1.43 mm. Middle leg Femur Hind leg Femur Male ices. opines eeaieanretioss 25.47 mm. Male a2. 2 eee 33.81 mm. Females ia hes eee 19.19 mm. Female.) 23323 eae 21.67 mm. The measurements of the types of P. canace Schmidt are as fol- lows: Male (apterous): Length without genital segments 11.76 mm.; width of head 2.38 mm.; width of pronotum 2.57mm.; length of mesonotum 3.19 mm.; greatest width of body across mesoacetabula 3.90 mm. Female (apterous). Length 13.0mm.; width of head 2.05 mm.; width of pronotum 2.05 mm.; length of mesonotum 2.86 mm.; great- est width of body across mesoacetabula 3.67 mm. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment 1 Ie Vath ee eee tte meee weeny ary 9.19 mm. 7.67 mm. 4.76 mm. Bemalecevewssy su) woe ar 7.14 mm. 6.19 mm. 2.95 mm. 1.43 mm. Middle leg Femur Hind leg Femur Mall sah Sra aoe pci 24.76 mm. Males S24 -oi ae 31.90 mm. Hemales= sas = _. 19.05 mm. Hemale an. eee 22.38 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::80:77. Suranal plate with median lobe broad, surpassing lateral wings caudally;. lateral wings short, their anterior and posterior margins not parallel. Pygofer with its dorsolateral pro- jection stout, slightly before its middle, and about as long as lateral lobes of the suranal plates as seen from above; apical margin of pygofer broad, truncate; lateral margins of pygofer beyond lateral projections parallel or nearly parallel as seen from below; as seen from the side ventral line of pygofer medially curved upward, mak- ing a constriction. Paramere as seen from above with most of its shaft hidden by median lobe of suranal plate, its arm turned laterad and its tip spatulate and broad. Distal part of dorsum of first genital segment but slightly elevated as seen in side view, its venter transversely constricted. Female. Length of thorax:Abdomen::114:135. Seventh abdomi- nal tergite much shorter than sixth. Connexival spine long, arising Genus PTmOMERA AMYOT AND SERVILLE 431 from base of connexivum of seventh segment, turned medially and then caudally as a long spine. Dorsolateral lobe of seventh ab- dominal segment long, its tip plainly surpassing ninth abdominal tergite when visible. Upper margin of dorsolateral lobe of seventh abdominal segment definitely curved. In side view appearing short, without or with only the slightest indication of ventrolateral lobe. Comparative notes: The males of this species have the apical mar- gin of the pygofer broad as do P. chinai and P. timorensis from which they are separated as shown in the key. The females have the base of connexival spine arising from beneath the dorsal edge of the dorsolateral lobe of the seventh segment at its base and di- rected medially then caudally. This is true of P. agriodes Schmidt and P. laticaudata Hardwicke from which they are separated as shown in our key couplet 31. Types: The type of P. cingalensis Stal is in the Riksmuseet at Stockholm, Sweden. The types of P. canace Schmidt, one male and three females are now in the Polish Academy of Sciences in Warsaw. Distribution: The type of both P. cingalensis Stal and P. canace Schmidt came from Ceylon without a definite locality. We have before use the following: “Ceylon, Pundaluoya Ap. ’98 (from rocky stream)” 1 ¢ ; “Ceylon, Pundaluoya May ’97 (from rocky stream)” 1 @ ; “Ceylon, Pundaluoya, Aug. ‘97 (from rocky stream)” 4¢ ¢, 39 9. The above is from the Kirkaldy Collection, University of Kansas. “Ceylon, Suduganga River, Matale R. A. Senior-White” 1¢ (Torre-Bueno Collection, K. U.); “Ceylon, Pitamba Ela Jan. 26, 1958, K.L. A. Perera” 1 g 29 9; “Ceylon, Nindoma, Jan. 27, 1958 K. L. A. Perera” 39 9 (K. U.). Ptilomera (Ptilomera) agriodes Schmidt Plate VII 1926. Ptilomera agriodes Schmidt, Ent. Mitt. 15(1):63-4. 1927. Ptilomera agriodes, Esaki, Eos Rev. Esp. Ent. 3(3):259. 1933. Ptilomera agriodes, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:371, 373, 421. 1933. Ptilomera lachne Schmidt, Lundblad, Arch. Hydrobiol Suppl. 12:372, 373, 419 (Said is female of P. agriodes Schmidt ). 1960. Ptilomera agriodes, Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 537 fig. 612,613 (wings) fig. 618 (hind tarsus showing two segments that are fused ). Since P. lachne Schmidt is a synonym, the following references also go here. 1926. Ptilomera lachne Schmidt, Ent. Mitt. 15(1):64 (Described female from Tranquebar, S. India). 1927. Ptilomera lachne, Esaki, Eos Rev. Esp. Ent. 3(3):259. 1960. Ptilomera lachne Matsuda, Univ. Kansas Sci. Bull. 41(2):267. 432 THe UNIVERSITY SCIENCE BULLETIN PLATE VII Ptilomera (Ptilomera) agriodes Schmidt Fic. 1. Female from Anamalai Hills, S. India. Dorsal view. Fic. 2. Male from S. India, ventral view of genital segments. Note the large elongate dorsolateral projections of the pygofer. Fic. 3. Caudal portion of pygofer of above male seen from the left side. Fic. 4. Dorsal view of above male. Note the shape of suranal plate, the dorsally turned up tip of pygofer, the huge dorsolateral projections of pygofer and the more slender curved parameres. Fic. 5. Female of figure 1. Ventral view of sixth and seventh abdominal segments. This does not show the true shape of the distal median lobe of the seventh ventrite. Fic. 6. Male type of P. agriodes. Drawing made of the second genital seg- ment in a slide mount. Fic. 7. Female of figure 1. Abdominal end as seen from the rear to show the true shape of the distal median lobe of the seventh ventrite. Fic. 8. The seventh abdominal segment of P. lachne Schmidt type, drawn from a slide showing right side. Fic. 9. Shows the sixth and seventh abdominal segments of a female of P. agriodes Schmidt seen from left side. Genus PrILOMERA AMYOT AND SERVILLE 433 PLATE VII 434 THE UNIVERSITY SCIENCE BULLETIN Redescription based on an apterous male type from Trichinopoli, South India, and on apterous female type of P. lachne Schmidt from Tranquebar, S. India. Size: Male (apterous): Length 17 mm.*; width of head 2.57mm.; width of pronotum 3.1 mm.; greatest width of body across meso- acetabula 5.0 mm. Female (apterous). Length 14.00 mm.; width of head 2.33 mm.; width of pronotum 2.52 mm.; length of mesonotum 3.33 mm.; great- est width of body across mesoacetabula 4.52 mm. Color: Quite typical for the genus. Ground color brown with usual black markings. Side of thorax with a black longitudinal band that is divided by a reddish yellow undulate line leading for- ward from mesothoracic spiracle.| Venter covered by a silvery pile. Antennae black, with last two segments more or less covered by a silvery pile. Legs dark brown. Front femur with a black longitudi- nal band on each side, its venter dark brown but more or less covered with a silvery pile. Tibia and tarsus nearly black but covered also with a silvery pile. Middle and hind legs dark brown, with distal end of tibia and tarsus light brown and covered with silvery pile. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::175:40:50:36.4. Female: 1st:2nd:3rd:4th::144:30:40:31. Relative Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Mallee aie cae inc ea 9.52 mm. 8.24 mm. 5.00 mm. 2.05 mm. Female. e305. ee 8.19 mm. 6.81 mm. 3.95 mm. 1.76 mm. Middle leg ~ Femur Hind leg Femur Mailers iiearetert i ean ce lt 24.29 mm. Males ine Se ea eae 29.76 mm. iemmalle Siete eee ha) laces 21.33 mm. Remale 2) See 23.81 mm. The front femur of male rather stout, 1.47 mm. in diameter, with usual incised place on under side of distal end that is smooth and shiny, with two black shiny transverse protuberances near its middle and at its base a conical less prominent tubercle. Front tibia also has at its inner base a similar incised area with three black shiny elevations, it is rather stout, as wide as middle femur (0.76 mm.); tarsus much more slender, outer lobe of terminal cleft of second tarsal segment much shorter than inner lobe. Middle femur stouter but shorter than hind femur. More than distal half of middle femur * Length as given by Schmidt in his description. We found that the genital segments of all types had been removed and cleared and mounted in balsam. + This yellow line is usually obscured by a band of silvery hairs. Genus PriLOMERA AMYOT AND SERVILLE 435 with a dense brush of long hairs; basal part of middle femur with short inconspicuous hairs. Metasternum of male short, subequal to first two visible abdominal ventrites * together; of female meta- sternum subequal to first visible abdominal ventrite. Male. Length of pregenital segments:Genital segments: :102:70. Suranal plate with median lobe broadly rounded, its caudolateral margins undulate and extending caudally far beyond short lateral wings, lateral margins of which are nearly parallel, with caudolateral corner of wings slightly protruded. Pygofer with dorsolateral pro- jections that surpass lateral wings of suranal plate by half their length, their tips pointed and turned caudally, caudal end of pygofer as seen from above broadly rounded, as seen from side or rear turned up and pointed. Paramere long, its arm curved later- ally and its tip pointed and turned forward. Female. Length of thorax:Abdomen::136:176. Seventh ab- dominal tergite a little longer than sixth tergite. Connexival spine arising beneath incured edge of connexivum, turned medially then caudally, its tip reaching caudal end of dorsolateral lobe that is short, broad and directed obliquely downward. Ventrolateral lobe very slightly developed. Distal lobe of seventh sternite turned dorsally and exposed, with a median longitudinal carina, its lateral margins slightly converging caudally and its tip or distal edge transverse and slightly concave as seen from rear. Comparative notes: The long dorsolateral projections of pygofer separate this species from all others. The female has the base of the connexival spine arising from beneath the connexival edge of the seventh abdominal segment near its base. The dorsolateral lobe of seventh abdominal segment short, its tip only slightly surpassing ninth tergite when visible, and this separates it from P. cingalensis which has the dorsolateral lobe of seventh abdominal segment long. Types: The type of P. agriodes Schmidt is an apterous male from Trichinopoli, South India and the type of P. lachne Schmidt is an apterous female from Tranquebar. Both are now in the Polish Academy of Sciences in Warsaw, Poland. Distribution: We have both males and females from a single place in South India that are like types. Besides the type we have seen the following: “South India Feb. 1955 P. S. Nathan” “Shovaroy Hill, Nagalur 4000 ft.” 59 9 (deilated) K. U. Coll.; “South India Feb. 1955 P. S. Nathan” “Shovaroy Hill, Yorcaud 4500 ft.” 1g de- alated), 8 apterous g¢ g¢, 9 macropterous 9 ¢ (K. U. Coll.); * The first abdominal ventrite is absent in this genus. 436 THE UNIVERSITY SCIENCE BULLETIN “South India Anamalai Hills, Cinchona 3500 ft. Apr. 1957 P. S. Nathan” 4 macropterous ¢ ¢, 4 macropterous ? 9; 92% 2g, 18° 9 (dealated); 3g g and13¢ ¢ (apterous) (K. U. Coll. From the same locality in May 1957, 1 9 (dedlated) (K. U. Coll. “India Coimbatore Dist. Nilgiri Hill, Kollor Susai Nathan” 1 g 1 9 (apterous) (Mus. Nat. Hist. Basel, Switzerland). \. ); ? Ptilomera (Ptilomera) timorensis n. sp. Plate VIII Size: Male (apterous): Length 15.1 mm.; width of head 2.31 mm.; width of pronotum 2.57 mm.; length of mesonotum 3.95 mm.; great- est width of body across mesoacetabula 4.00mm. Female (apter- ous). Length 14.05 mm.; width of head 2.1 mm.; width of pronotum 2.57 mm.; length of mesonotum 3.95 mm.; greatest width of body across mesoacetabula 4.00 mm. Color: Only head and pronotum with the typical ground color of reddish brown; Meso- and metonotum dark brown to black; first six abdominal tergites black, seventh may be reddish brown; genital segments paler; in all of the specimens the silvery pile bands and spots are confined to the sides. The venter with the usual short silvery pile. The typical pattern of black markings not distinguish- able. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::153:35:42:31. Female: Ist:2nd:3rd:4th::120:28:35:27. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg - Femur Tibia segment segment Male (paratype) ....... 7.77 mm. 6.81 mm. 3.71 mm. 1.57 mm. Female (paratype) ..... 6.57 mm. 5.48 mm. 3.33 mm. 1.38 mm. Middle leg Femur Hind leg Femur Male (paratype) ........ 22.14 mm. Female (paratype) ..... 17.14 mm. Male (paratype) ........ 27.14 mm. Female (paratype) .... 19.05 mm. Male. Length of pregenital abdominal segments:Genital seg- ments: :86:84. Genital segments: Suranal plate with median lobe not broad but surpassing caudally the short broad lateral wings, the anterior and posterior margins of which are not parallel. Pygofer of unusual shape, quite uniformly broad, its apical margin broad with a median incission or notch, dorsolateral projections of pygofer short and slender. Shaft of paramere rather slender but almost entirely hidden by the median lobe of suranal plate. The arm is long. Fic. Fic. Fic. Fic. Fic. ments. Fic. = OTS 99 bo > GENus PTILOMERA AMYOT AND SERVILLE 437 PLATE VIII \V/ 5 Ptilomera (Ptilomera) timorensis n. sp. Female allotype, dorsal view. Male holotype, ventral view of genital segments. Male holotype, dorsal view. Male holotype, arm of left paramere. Female allotype, ventral view of sixth and seventh abdominal seg- Female allotype, side view of sixth and seventh abdominal segments. 438 THE UNIVERSITY SCIENCE BULLETIN Female. Length of thorax:Abdomen::110:128. Seventh abdomi- nal tergite a little longer than the sixth. Connexival spines very short, inconspicuous, arising from end of connexivum and turned medially but too short for their tips to meet. Dorsolateral lobe short and broad at base without or with only the faintest indica- tion of a ventrolateral lobe of seventh abdominal segment. Distal lobe of seventh abdominal sternite entirely exposed and not turned dorsally, its lateral margins nearly straight but converging, distal end slightly rounded as shown on Plate VIII fig. 5. Comparative notes: This striking species is so different from any other that it can be recognized by anyone. The males by the broad pygofer with its distal margin at least slightly bi-lobed and the female by its very short connexival spines. Types: This apterous type series consists of holotype male bear- ing the label, “Timor, Baaguia, August 1935. C. Buhler u. Meyer.” Allotype with the label, “Timor Soe. June 1, 1935 C. Buhler u. Meyer.” Paratypes from Timor Soe 2 ¢ g,and1 ¢ labeled “Timor.” Kirkaldy Coll. in K. U. Holotype allotype and 1 g¢ and 19 para- types are in Basel Museum, Switzerland and one pair in K. U. col- lection. Distribution: Known only by the type localities. Ptilomera (Ptilomera) chinai n. sp. Plate IX Size: Male (apterous): Length 17.14mm.; width of head 2.38 mm.; width of pronotum 2.52mm.; length of mesonotum 3.57 mm.; greatest width of body across mesoacetabula 3.90 mm. Female (apterous). Length 17.62 mm.; width of head 2.24 mm.; width of pronotum 2.29 mm.; length of mesonotum 3.52 mm.; great- est width of body across mesoacetabula 4.14 mm. Color: Pattern typical for the genus. In these two specimens more of the ground color is exposed because the black areas are small. Vertex with two converging black stripes. The neck crossed by a black band; anterior margin of pronotum with a faint, dark band; two narrow longitudinal black lines on side of the meso- and metathorax separated by a broader band of ground color, which is not densely covered by silvery pile. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::164:38:53: ?. Female: Ist:2nd:3rd:4th::145:32: P: ?, Genus PTILOMERA AMYOT AND SERVILLE 439 PLATE Ix Ptilomera (Ptilomera) chinai n. sp. Fic. 1. Ventral view of genital segments of male holotype. Fic. 2. Apterous male holotype. Fic. 3. Apterous female allotype. Fic. 4. Ventral view of sixth and seventh abdominal segments of female allotype. Fic. 5. Female allotype, left side view of sixth and seventh abdominal seg- 440 THE UNIVERSITY SCIENCE BULLETIN Actual Lengths of Leg Segments 1st tarsal 2nd _ tarsal Front leg Femur Tibia segment segment Male (holotype) ....... 8.67 mm. 7.33 mm. 4.48 mm. Female (allotype) ...... 8.05 mm. 7.10 mm. 6.76 mm. 1.62 mm. Middle leg Femur Hind leg Femur Male (holotype) ........ 25.95 mm. Male (holotype) ...... 28.60 mm. Female (allotype) ....... 22.76 mm. Female (allotype) .... 26.19 mm. Male (holotype). Length of pregenital abdominal segments: Genital segments: :92:89. Genital segments: Median lobe of suranal plate moderately broad; lateral wings well developed, thickened to- ward their ends; their front and rear margins not parallel because the rear margin of basal half of the lateral wing of suranal plate concave, the wings surpass caudally the median lobe. Pygofer, as seen from below, with its lateral margins beyond the dorsolateral projections parallel or nearly so. Pygofer truncate (or slightly medially concave) at tip. Dorsolateral projections of pygofer as seen from above short, a little longer than the lateral wings of the suranal plate and rather pointed. Paramere long, the exposed shaft short, the long arm curved laterally, its blunt tip slightly turned caudally. The first genital segment with a distinct median longi- tudinal keel on its venter. Female. Length of thorax:Abdomen::135:192. Seventh abdomi- nal tergite only slightly shorter than sixth tergite. Moderately long connexival spines arise at end of connexivum. Dorsolateral lobe long, its base a long triangular lobe that becomes cylindrical, slightly sigmoid on upper margin and spine like. Ventrolateral lobe absent or very faintly indicated. Sides of the usually vertically directed distal lobe of the seventh abdominal sternite nearly straight and nearly parallel. Comparative notes: Of the five species described from Borneo this is the only one in which the males have a pygofer with a broad nearly truncate tip and the females with a long dorsolateral lobe that is slightly sigmoid on upper margin, as seen laterally. Types: Male holotype and female allotype both apterous, bear the label “Sambas Borneo (West Coast).” “C. J. Brooks, B. M. 1936-681.” These are in the British Museum. Two paratypes, also apterous, are in the Bernice Bishop Museum in Hawaii. They bear the label “Borneo: Sarawak Kuching, Santubong 797 to 1500 m. June, 18-30, 1958. T. C. Maa, Collector No. M. B.-78"13,1¢. Distribution: Known only from the type localities. This species is named in honor of Dr. W. E. China of the British Museum. Genus PrmrOMERA AMYOT AND SERVILLE 44] Ptilomera (Ptilomera) nunikanensis n. sp. Plate X Size: Male holotype (apterous): Length 14.52 mm.; width of head 2.19 mm.; width of pronotum 2.33 mm.; length of mesonotum 3.24mm.; greatest width of body across mesoacetabula 3.43 mm.? Female allotype (apterous): Length 15.48 mm.; width of head 2.05mm.; width of pronotum 2.05mm.; length of mesonotum 3.23 mm.; greatest width of body across mesoacetabula 3.57 mm.? Color: Both specimens are teneral and the ground color nearly testaceous and black areas reduced. However, the pattern is typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::152:34:48:31. Female: Ist:2nd:3rd:4th::145:30.5:?:?. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Males eames ene eae a, 8.10 mm. 6.57 mm. 3.95 mm. 1.57 mm. eran all ewe eee eee ener ee: 7.62 mm. 6.29 mm. 4.00 mm. 1.57 mm. Middle leg Femur Hind leg Femur Nia errs ieee or a 21.52 mm. IN Teall aa ees sais be entre vane 25.47 mm. Hemalen ea i ecu e ae 20.00 mm. Hemial en aa oe ti 21.91 mm. Male holotype. Length of pregenital abdominal segments:Genital segments: :82:82. Genital segments: Suranal plate with median lobe small, lateral wings well developed and plainly surpassing median lobe caudally. Pygofer as seen from below conspicuously broad beyond the middle. Caudal half of that section of pygofer beyond its dorsolateral projections parallel sided. As seen laterally this distal portion is turned dorsally to a truncate tip as shown on plate X fig. 7. Dorsolateral projection of pygofer stout and extends laterally farther than the wing of suranal plate. The paramere stout, its shaft short, its arm long and of characteristic shape as shown on plate X fig. 4. Female. Length of thorax: Abdomen: :125:158. Seventh abdomi- nal tergite about as long as sixth tergite. Connexival spine long, arising at end of connexivum. Dorsolateral and ventrolateral lobes separated by a narrow incision. Ventrolateral lobe sharp pointed. Comparative notes: This species is quite distinct from other species of Ptilomera from Borneo. Types: Holotype (apterous male) and allotype (apterous fe- male) bear the labels “Nunikan I. E. Kalimantan Dist. Indonesian 442 Fic. Fic. Fic. Fic. Fic. ments. Fic. Fic. CON CoE e THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) nunikanensis n. sp. Male holotype, dorsal view. Female allotype, ventral view of sixth and seventh segments. Female allotype, dorsal view. Enlarged drawing of dorsal view of right paramere. (See fig. 1.) Female allotype, left side view of sixth and seventh abdominal seg- Male holotype, ventral view of genital segments. Male holotype, right side view of genital segments. Genus PrmOMERA AMYOT AND SERVILLE PLATE X 443 444 THe UNIVERSITY SCIENCE BULLETIN Borneo, Forest stream XII 1953, Werner von Hentig leg.” Both holotype and allotype are in the Chicago Mus. Nat. Hist. Distribution: Known only from the type locality. Ptilomera (Ptilomera) kirkaldyi n. sp. Plate XI We describe below two males from “Ost Borneo” as this species. The female came from “West Borneo” and may prove to be some other species. Therefore, we do not consider it as an allotype. Size: Male holotype (apterous): Length 19.05 mm.; width of head 2.52 mm.; width of pronotum 2.90 mm.; length of mesonotum 4.33 mm.; greatest width of body across mesoacetabula 4.43 mm. Male paratype (macropterous ). Length including wings 17.62 mm.; width of head 2.19mm.; width across humeri 3.33 mm.; greatest width of body across mesoacetabula 3.88 mm. Female from West Borneo (apterous): Length 17.14 mm.; width of head 2.19mm.; width of pronotum 2.19mm.; length of meso- notum 3.43mm.; greatest width of body across mesoacetabula 4.10 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::190:43:?:?. Female: Ist:2nd:3rd:4th: :147:33:?:?. Actual Lengths of Leg Segments Ist tarsal Qnd tarsal Front leg Femur Tibia segment segment Male (holotype) ....... 10.00 mm. 8.67 mm. 5.24 mm. 2.00 mm. Female (W. Borneo) .... 8.09 mm. 6.76 mm. 4.19 mm. Middle leg Femur Hind leg Femur Male (holotype) ........ 29.1 mm. Male (holotype) ...... 34.76 mm. Female (W. Borneo) .... 22.05 mm. Female (W. Borneo) .. 25.71 mm. Male holotype. Length of pregenital abdominal segments:Genital segments::98:87. Genital segments: Suranal plate with median lobe short, plainly surpassed caudally by the broad lateral lobes. Pygofer, as seen from below conspicuously broad beyond its middle, the sides converging to a dorsally directed point. In side view the pygofer has a slight transverse constriction. The paramere shaped as shown on Plate XI, fig. 7. Venter of first genital segment with a median longitudinal keel. Female. This unattached female from West Borneo is unlike that of any other Borneo species so we tentatively place it here. Length of thorax:Abdomen::132:184. Seventh abdominal tergite Genus PrmomerA AMYOr AND SERVILLE 445 PLATE XI en, Ptilomera (Ptilomera) kirkaldyi n. sp. Fic. 1. Male macropterous paratype from Borneo, Ost. Fic. 2. Apterous female from Nowong, W. Borneo of P. kirkaldyi (?). Fic. 3. Abdominal venter of apterous female from Nowong, W. Borneo of P. kirkaldyi (?). Fic. 4. Holotype apterous male from Borneo, Ost. Fic. 5. Apterous female from Nowong, W. Borneo. Side view of sixth and seventh abdominal segments of P. kirkaldyi (?). Fic. 6. Venter of genital segments of male holotype. Fic. 7. Exposed portion of shaft and arm of left paramere of holotype. 446 THE UNIVERSITY SCIENCE BULLETIN a little shorter than the sixth. Connexival spine long, arising from end of connexivum. Dorsolateral lobe of seventh abdominal seg- ment long, spine like for its entire length as seen laterally. Ventro- lateral lobe absent or but faintly indicated. Sides of the vertically directed distal lobe of the seventh abdominal sternite straight and nearly parallel. Comparative notes: The closest relatives of this species are from other species from Borneo. Types: Holotype apterous male and a paratype macropterous male are from East Borneo. They are in the K. U. Kirkaldy Collec- tion. The female we tentatively assign to this species bears the label, “Mowong, W. Borneo F. Muir, Sept. 1907” and is in the Cali- fornia Academy of Sciences. It is not a type. Distribution: Known only from above. Ptilomera (Ptilomera) gressitti n. sp. Plate XII Size: Male holotype (apterous): Length 17.61 mm.; width of head 2.45 mm.; width of pronotum 2.24 mm.; length of mesonotum 3.66 mm.; greatest width of body across mesoacetabula 4.19 mm. Female paratype (apterous). Length 17.38 mm.; width of head 2.21mm.; width of pronotum 2.43mm.; length of mesonotum 3.81 mm.; greatest width of body across mesoacetabula 4.38 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male holotype: Ist:2nd:3rd:4th::173:41: ?: ?. Male paratype: Ist:2nd:3rd:4th: :162:40:50:34. Female paratype: Ist:2nd:3rd:4th::158:37:48:28--. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Male holotype =. ..:. 5... 9.52 mm. 7.81 mm. 4.86 mm, 2.00 mm. Female paratype ....... 6.24 mm. 7.24 mm. 4.47 mm. 1.76 mm. Middle leg Femur Hind leg Femur Male holotype .......... 26.43 mm. Male holotype ....... 30.95 mm. Female paratype ........ 22.85 mm. Female paratype ...... 25.00 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::87:72. Genital segments: Median lobe of suranal plate small; the short broad lateral wings only slightly surpassing median lobe, their rear margin not longer than the width of their tips. Pygofer as seen from below unusually broad beyond its lateral pro- Genus PTm@OMERA AMYOT AND SERVILLE 447 PLATE XII Ptilomera (Ptilomera) gressitti n. sp. Fic. 1. Female, dorsal view. Fic. 2. Male, ventral view of genital segments. Note the long parameres and the unusual width of pygofer at the caudal base of the lateral projections. Fic. 3. Male, dorsal view. Fic. 4. Female, ventral view of sixth and seventh abdominal segments. Note the very large median caudal lobe of seventh ventrite. Fic. 5. Female, side view of above. Note that the median caudal lobe is not turned dorsad as it is in most species. 448 THe UNIVERSITY SCIENCE BULLETIN jections which are slender, pointed and only slightly surpassing laterally the broad short lateral wings of suranal plate. Distal por- tion of pygofer with sides converging, its tip bluntly rounded. Para- mere long, shaft and arm inseparable, as shown on Plate XII fig. 2. Female. The length of thorax:Abdomen::142:172. Seventh ab- dominal tergite as long as the sixth. Connexival spine moderately long and arising from end of connexivum. In side view the species is unlike any other species. At first glance the dorsolateral lobe appears to have a notch on its lower margin and no ventrolateral lobe. However, the lower lobe beyond the notch is the ventrolateral lobe and therefore in the key it runs out to couplet 36. The distal lobe of seventh abdominal sternite is unusually large and not turned vertically. Comparative notes: The female is unique, and its closest relative may be P. nunikanensis. The male, however, appears to be related to P. kirkaldyi and P. nunikanensis. Types: Described from an apterous series containing 3 ¢ ¢ , and 29 9 that bear the following labels: “British N. Borneo, West Coast Residency, Ranau 8 mi. N. Paring Hot Springs, 500m. Oct. 8-11, 1958” “T. C. Maa collector” Holotype, allotype and one male paratype are in the Bernice P. Bishop Museum, Honolulu, Hawaii, and one pair of paratypes are in the K. U. collection. Distribution: Known only from the type locality. Ptilomera (Ptilomera) hylactor Breddin Plate XIil 1903. Ptilomera hylactor Breddin, Societas Entomologica 17(19):148. (De- scribed from annam. @. and @ ). 1927. Ptilomera hylactor, Esaki, Eos, Rev. Esp. Ent. 3(3):259. 1933. Ptilomera hylactor, Hoffmann, Lingnan Sci. Jour. 12 Suppl. p. 247. 1933. Ptilomera hylactor, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 419. 1941. Ptilomera hylactor, Hoffmann, Lingnan Sci. Jour. 20(1):35. 1960. Eee hylactor, Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 539, fig. Of the ten species described by Breddin, he gave one or more figures for six of them. He gave no figure for P. hylactor. How- ever, his description is adequate to recognize this striking species. He said it was the largest species known to him but it would have been more accurate to say it is the longest and most slender species. For the redescription of this species we have before us three speci- mens from Breddin’s collection that bear the label, “Annam Phuc- Son Nov.-Dez. H. Fruhstorfer,” one male without genitalia, one fe- male and one nymph. These now bear the red label “Typus” “det. Genus PrmoMERA AMYOT AND SERVILLE 449 PLATE XIII Ptilomera (Ptilomera) hylactor Breddin Fic. 1. Female from type series, dorsal view. Fic. 2. Ventral view of sixth and seventh abdominal segments of above fe- male. Fic. 3. Male from type series, ventral view of genital segments. Note the broad complete median keel on first genital segment and remarkable shape of the venter of the pygofer. Fic. 4. Male from type series, dorsal view. Fic. 5. Female, rear view of middle lobe of seventh abdominal ventrite. Fic. 6. Female, sixth and seventh abdominal segments as seen from left side. 450 THE UNIVERSITY SCIENCE BULLETIN by Lundblad.” We find in our Torre-Bueno Collection a pair with the same identical label as above and these enable us to give the following information: Size: Breddin gave the length 18.5—19.5 mm. Hind femur and trochantur ¢ 32.5mm. 9@ 26.5mm.; while our Torre-Bueno pair are as follows: Male (apterous): Length 18.1 mm.; width of head 2.17 mm.; width of prothorax 2.24 mm.; length of mesonotum 3.57 mm.; great- est width of body across mesoacetabula 3.64 mm. Female (apterous): Length 18.57 mm.; width of head 2.12 mm.; width of pronotum 2.24 mm.; length of mesonotum 3.71 mm.; great- est width of body across mesoacetabula 4.00 mm. Color: Quite typical for the genus. However, the four adults available to us have the abdominal tergites black. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::158:40:?:?. Female: Ist:2nd:3rd:4th: :162:33:?:?. Actual Lengths of Leg Segments lst tarsal 2nd tarsal Front leg Femur Tibia segment segment Nia etee er eek thorn eet. 8.81 mm. 7.23 mm. 4.38 mm. 1.85 mm. Hem~aley tons we ee 8.19 mm. 6.57 mm. 4.33 mm. 1.91 mm. Middle leg Femur Hind leg Femur INT alll Se se area Dent 24.52 mm. Maley 22.3) ae eerie 29.05 mm. Bemlalles ee eer eee. 21.9 mm. Femalesis32 eo eee 25.00 mm. Male. Length of pregenital segments:Genital segments: :102:85. Genital segments: Suranal plate with median lobe very short but moderately broad. Lateral wings of suranal plate short, moderately broad, their front and rear margins not parallel and not surpassing median lobe. Venter of pygofer with basal two thirds narrow, re- sembling a broad keel, sides above the keel concave to a position beyond the lateral projections of pygofer. As seen from below pygofer not conspicuously broad beyond its middle. Apex of pygofer bluntly rounded. Venter of first genital segment with a well de- veloped keel. In dorsal view first genital is plainly longer than seventh tergite or than suranal plate. Female. Length of thorax:Abdomen::142:207. Seventh abdomi- nal tergite as long as sixth. Connexival spine moderately long, arising from the end of the connexivum. Incission between dorso- lateral and ventrolateral lobes broad and shallow. Both lobes long and spine-like. Dorsolateral lobe of seventh abdominal segment long, stout and straight, its upper margin from base of connexival Genus PrmOMERA AMYOT AND SERVILLE 451 spine to its tip nearly twice as long as connexival margin of seventh abdominal segment. Distal lobe of the seventh sternite rather long and narrow, turned dorsally and embraced by lateral lobes. Comparative notes: This species stands quite alone, but runs out to couplet 13 in the key to males, and to couplet 47 in key to females. Types: An apterous male without the genital segments, and an apterous female and one nymph, all bearing the label “Annam Phuc- Son Noy. Dez. H. Fruhstorfer.” All types are in Deutsches En- tomologisches Institut, Berlin. Distribution: Known only from the type series, a pair of which we found undetermined in our Kirkaldy Collection at K. U., and 1 2 bearing the label “Haut Mekong Pau Lan 13 V. 1918. R. V. de Salvaza” (Brit. Mus.). Ptilomera (Ptilomera) breddini n. sp. Plate XIV Size: Male paratype (apterous): Length 15.62mm.; width of head 2.1 mm.; width of pronotum 2.36 mm.; length of mesonotum 3.14 mm.; greatest width of body across mesoacetabula 3.48 mm. Female paratype (apterous): Length 15.00 mm.; width of head 2.1mm.; width of pronotum 2.14mm.; length of mesonotum 2.90 mm.; greatest width of body across mesoacetabula 3.86 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::142:38:?:?. Female: Ist:2nd:3rd:4th::128:28:?:?. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Male aaron 7.61 mm. 6.67 mm. 3.71 mm. 1.43 mm, Hemaleinc 25h scoses oe 6.67 mm. 5.81 mm. 3.81 mm. 1.43 mm. Middle leg Femur Hind leg Femur Male ety paral eee 20.48 mm. Males siete cine ete 22.62 mm. em al ent tas nee eo 18.57 mm. Female: (geo. c ie n 19.52 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::91:81. Genital segments: Suranal plate with median lobe broader than the lateral wings which are long, surpassing caudally the median lobe and have their front and rear margins nearly parallel. Pygofer with its dorsolateral projections short and rather slender. Pygofer as seen from below not conspicuously broad be- yond its middle and appearing rather elongate, its venter appear- Fic. Fic. Fic. Fic. Fic. Fic. ments. Fic. ments. DU © pop THe UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) breddini n. sp. Female allotype, dorsal view. Male holotype, ventral view of genital segments. Male holotype, dorsal view. Male holotype, side view of left paramere. Male holotype, rear view of arm of left paramere. Female allotype, ventral view of sixth and seventh abdominal seg- Female allotype, left side view of sixth and seventh abdominal seg- Genus PrmoMERA AMYOT AND SERVILLE PLATE XIV 453 454 THe UNIVERSITY SCIENCE BULLETIN ing slender, its tip pointed. Paramere with its shaft longer than its arm. The arm as seen from the rear, with a ventral bulge or keel as shown of Plate XIV figs. 4, 5. Female. Length of thorax:Abdomen::116:160. Seventh abdomi- nal tergite about as long as sixth. Connexival spine moderately long and arising from end of seventh connexivum. Seventh ab- dominal segment with dorsolateral and ventrolateral lobes separa- rated by a deep incission; dorsolateral lobe rather slender, only moderately long, its upper margin from base of connexival spine to its tip much less than twice as long as connexival margin of seventh abdominal segment. Distal lobe of seventh abdominal sternite of usual size and embraced by the lateral lobes, its sides straight and nearly parallel and directed vertically. Comparative notes: Nearest relative is P. papuensis n. sp. which runs out with it to couplet 16 of the key to males and to couplet 49 in the key to females. Types: Male holotype, allotype and 2 males and 1 female para- types bearing the label “N. Guinea S. E. Haveri, Loria VII, XI-93.” and one male “Milne Bay, N. Guinea” one female “N. Guinea, S. E. Paunemu riv. Loria IX-XII 92”. Types are in the Kirkaldy Collec- tion at K. U. Distribution: Known only by the type localities. Ptilomera (Ptilomera) papuensis n. sp. Plate XV Size: Male paratype (apterous): Length 15.00 mm.; width of head 2.02 mm.; width of pronotum 2.42 mm.; length of mesonotum 3.19mm.; greatest width of body across mesoacetabula 3.67 mm. Female paratype (apterous). Length 14.52 mm.; width of head 1.95mm.; width of pronotum 2.24mm.; length of mesonotum 3.19 mm.; greatest width of body across mesoacetabula 3.81 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th: :138:35:49:31. Female: Ist:2nd:3rd:4th::125:29:40:28. Actual Lengths of Leg Segments Ast tarsal 2nd_ tarsal Front leg Femur Tibia segment segment Male paratype .......... 7.61 mm, 6.67 mm. 3.71 mm. 1.43 mm. Female paratype ........ 6.67 mm. 5.81 mm. 3.81 mm. 1.43 mm. Middle leg Femur Hind leg Femur Male paratype! =4.- 20.48 mm. Male paratype ......... 22.62 mm. Female paratype ........ 18.57 mm. Female paratype ...... 19.52 mm. GeNus Prmi.OMERA AMYOT AND SERVILLE 455 Male. Length of pregenital abdominal segments:Genital seg- ments: :80:118. Genital segments: Suranal plate with median lobe not large, not wider than width of lateral wings which are long, surpassing caudally the median lobe and have their front and rear margins nearly parallel. Pygofer with its dorsolateral projections short and rather slender. Pygofer as seen from below not con- spicuously broad beyond its middle and rather elongate, its venter appearing slender, its tip pointed. Paramere with shaft longer than its arm. See Plate XV fig. 4. Female. Length of thorax:Abdomen::135:163. Seventh tergite slightly shorter than sixth. Connexival spine moderately long and arising from end of seventh connexivum. Seventh abdominal seg- ment with dorsolateral and ventrolateral lobes separated by a deep incission; dorsolateral lobe stout, only moderately long, more or less spine-like, its upper margin from base of connexival spine to its tip much less than twice as long as connexival margin of seventh abdominal segment. Distal lobe of seventh abdominal sternite of usual size. Comparative notes: Its nearest relative is P. breddini n. sp., which runs out with it to couplet 16 of the key to males and to couplet 49 in the key to females. Types: Male holotype, allotype and 5 males and 5 female para- types, all apterous, bear the label “Mt. Lamington N. E. Papua 1300-1500 ft. C. T. McNamara.” One pair of paratypes are in the K. U. collection. All others are in S. Australian Museum. Distribution: Besides the above types there are three males and one female which bear the label “Finsch Haven, New Guinea L. Wagner.” These are very pale specimens and appear teneral. The males appear to be this species but the female has the ventrolateral lobe of the seventh segment broadly truncate at tip but may repre- sent the same kind of variation that appears to occur in P. harpyia Schmidt. 456 THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) papuensis n. sp. Fic. 1. Female from N. E. Papua, dorsal view. Fic. 2. Male from above place, ventral view of genital segments. Fic. 3. Male from above place, dorsal view. Note the long parallel sided lateral wings of suranal plate. Fic. 4. Male. Left paramere as seen from the side. Fic. 5. Same female as in figure 1, ventral view of sixth and seventh ab- dominal segments. Fic. 6. Female, side view of sixth and seventh abdominal segments. Genus PritoMERA AMYOT AND SERVILLE PLATE XV 457 458 THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) aéllo aéllo Breddin Plate XVI, figs. 3, 8 and 9 1906. Ptilomera aéllo Breddin, Societas Entomologica, 21(2):9. (Described from 3 apterous 2 2, 1 nymph from “Timmena, Nieuw Guinea Ex- peditie 1903”). 1958. Ptilomera aéllo Hungerford and Matsuda, Bull. Brooklyn Ent. Soc. 53(3):69. 1960. Ptilomera aéllo Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 539. This species from New Guinea has been overlooked by other writers for fifty years. It was published in an obscure journal which claimed to be the “Organ of the International Entomological So- ciety” and we found it in R. 9 of the bound volumes of Kirkaldy’s library of separates in the University of Kansas. In his description Breddin wrote that this species is broader and shorter than the other species known to him, the abdomen shorter in relation to the thorax. His color description is not specific. He gave no illustra- tion but said that the contour of the side plate of the seventh ad- dominal segment was nearest to that of P. pamphagus Breddin. We therefore give the following redescription and figures of his types: Size: Breddin gave the length as 14.00 mm. including the con- nexival spines 145mm. We here give the measurements of the largest of the three apterous female cotypes. Length 14.00 mm.; width of head 2.05mm.; width of pronotum 2.19mm.; length of mesonotum 3.09 mm.; greatest width of body across the mesoace- tabula 4.33 mm. Color: Typical for the genus. The ground color and black mark- ings subject to the same variations that occur in most other species. For example, one cotype has four distinct black dots on the vertex, as described by Breddin, a second has front and rear dots on each side connected by an embrowned line, and the third cotype has the embrowned connecting line nearly black so that the vertex has two converging lines on the vertex instead of four dots. Structural characteristics: Relative lengths of antennal segments. Ist:2nd:3rd:4th: :125:33:46:31. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Female type: ee 6.9 mm. 5.9 mm 3.81 mm. 1.38 mm. Middle leg Femur Hind leg Femur Hemale types. c eh. 19.14 mm. Female type .......... 21.19 mm. Female. Length of the thorax:Abdomen::114:132 in one cotype and ::120:135 in another. Thus this is not a character that would Genus PTmOMERA AMYOT AND SERVILLE 459 separate this species from others as Breddin thought. Since the male of this species is unknown, we can only give the characters of the seventh abdominal segment of the female. Seventh tergite a little shorter than sixth. Connexival spines long, arising from the curved edge of the connexivum. Dorsolateral and ventrolateral lobes short, separated by a moderately narrow and deep incission; dorsal margin from base of connexival spine to its tip nearly straight or slightly concave and shorter than the connexival spine. Comparative notes: This is one of species in which the incission between dorsolateral and ventrolateral lobes is relatively small in the female. Yet this species can be recognized from the couplet 46 in the key to female. Types: Three females from “Nieuw Guinea Expeditie Timmema 6-7 apr.” In Deutsches Entomologisches Institut in Berlin. Distribution: Known only from the type locality. Ptilomera (Ptilomera) aéllo cheesmanae n. subsp. Plate XVI, figs. 1, 2, 4, 5, 6, 7, 10 We have three species from New Guinea, none of which possesses females like the type of P. aéllo Breddin, but one is certainly close. Therefore we are placing it as a subspecies. Size: Male holotype (apterous): Length 16.67 mm.; width across head 2.48 mm.; width of pronotum 7.71 mm.; length of mesonotum 8.81 mm.; greatest width of body across mesoacetabula 4.43 mm. Female allotype (apterous): Length 15.24mm.; width across head 2.24 mm.; width of pronotum 2.38 mm.; length of mesonotum 3.14 mm.; greatest width of body across mesoacetabula 4.57 mm. Color: Typical for the genus. In most of the specimens the vertex has a black line along the inner margin of the eye that joins the front end of each of the two converging lines. Structural characteristics: Relative lengths of antennal segment- Male: Ist:2nd:3rd:4th: :167:44:54:36. Female: Ist:2nd:3rd:4th::135:35:47:31. Actual Lengths of Leg Segments Ist tarsal 2nd _ tarsal Front leg Femur Tibia segment segment IMalemeny wera. vi ge au . 8.95 mm. 6.76 mm. 5.14 mm. 1.71 mm. Hemialemtcun iste ewer 7.15 mm. 6.29 mm. 3.00 mm. 1.48 mm. Middle leg Femur Hind leg Femur Male is ier et trite cae 25.90 mm. Mia ete uiiiere Okan erate 30.71 mm. Bemale Gry etree. pe 20.48 mm. Hemalem ieee aes 22.48 mm. 460 Fic. Fic. ments. Fic. ments. Fic. Fic. Fic. Fic. Fic. Fic. Fic. THE UNIVERSITY SCIENCE BULLETIN PLATE XVI Ptilomera (Ptilomera) aéllo aéllo Breddin 3. Female cotype, dorsal view. 8. Female cotype, ventral view of sixth and seventh abdominal seg- 9. Female cotype, lateral view of sixth and seventh abdominal seg- Ptilomera (Ptilomera) aéllo cheesmanae n. subsp. Female, dorsal view. Female, ventral view of sixth and seventh abdominal segments. Male, dorsal view. Male, ventral view of genital segments. Male, left paramere. Sig Ohs Camis POE ea Male, dorsal view of second genital segment. 10. Female, lateral view of sixth and seventh abdominal segments. 18—1367 GeNus PriLoMERA AMYOT AND SERVILLE PEATE XVit 461 462 THe UNIVERSITY SCIENCE BULLETIN Male. Length of pregenital abdominal segments:Genital seg- ments::85:62. Genital segments: Suranal plate with median lobe not large. Lateral wings of suranal plate with front and rear mar- gins nearly parallel, not surpassing median lobe caudally. Lateral wings rather long, rear margins nearly twice as long as their width at their tips. Pygofer with short dorsolateral projections; Apex of pygofer pointed; Pygofer as seen from below not conspicuously broad beyond its middle; paramere as shown on Plate XVI figs. 6, 7. Female. Length of thorax:Abdomen::130:152. Seventh tergite considerably shorter than sixth. Connexival spines not as long as in the true P. aéllo. The connexival ridge convexly curved laterally before forming the base of connexival spine. Dorsolateral and ventrolateral lobes short, separated by a moderately narrow in- cission; dorsolateral lobe turned obliquely down, its dorsal margin from base of connexival spine to its tip convexly curved and longer than connexival spine. Comparative notes: The key will separate the female from that of the typical P. aello, and the male in couplet 17 from P. oribasus Breddin. Types: Described from apterous male holotype, apterous female allotype and two male and two female paratypes from “Dutch New Guinea, Cyclops Mts. Mt. Lina 3500 ft. HI 1936 L. E. Cheesman” and 3 male and 2 female paratypes from “Dutch New Guinea Cyclops Mts. Sabron. 930 ft. IV 1936 L. E. Cheesman. All are apterous. The type series is in the British Museum. 1 male and 1 female are in K. U. collection. Distribution: Known only for the type localities. Ptilomera (Ptilomera) tigrina Uhler Plate XVII, figs. 1, 2, 4 1860. Ptilomera tigrina Uhler, Pro. Acad. Nat. Sci. Philadelphia June 1860: 230. (Described from Hong Kong). 1866. Ptilomera tigrina, Mayr, Novara-Exped. Zool. Theil. Bd. If Abth. 1:177. (Gave this as synonym of P. laticaudata (Hardw.) in error). 1904. Ptilomera tigrina, Distant, Faun. Brit. Ind. Rhynch. II:185. (Gave this as synonym of P. laticaudata (Hardw.) in error). 1927. Ptilomera tigrina, Esaki, Eos, Rev. Esp. Ent. 3(3):259. (Listed). 1933. Ptilomera tigrina, Wu, Lingnan Sci. Jour. 12 Suppl:209. (Listed). 1933. Ptilomera tigrina, Hoffmann, Lingnan Sci. Jour. 12 Suppl.:247. (Listed). 1933. Ptilomera tigrina, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 418. (He did not see this species ). 1935. Ptilomera tigrina, Wu, Ins. Sinen. 2:543. (Catalogue). 1941. Ptilomera tigrina, Hoffmann, Lingnan Sci. Jour. 20(1):35. (Catalogue). Genus PriLtoOMERA AMYOT AND SERVILLE 463 For more than one hundred years this species has remained un- recognized although three female types have been in the U. S. National Museum in Washington D. C. during all this time. Redescription of these female types: Size: Female paratype (apterous). Length 16.43 mm.; width of head 2.05 mm.; width of pronotum 2.12 mm.; length of mesonotum 3.29 mm.; greatest width of body across mesoacetabula 3.71 mm. Color: Color and pattern typical for the genus. These types have the abdominal tergites mostly of the ground color with the black markings confined to the anterior and lateral margins of each tergite. Structural characteristics: Antennae are missing. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Female (paratype) ...... 7.00 mm. 6.05 mm. 3.67 mm. 1.48 mm. Middle leg Femur Hind leg Femur Female (paratype) ...... 20.14 mm. Female (paratype) .... 22.14mm. Males are unknown but we suspect that when they are found P. harpyia Schmidt will be a synonym of this species. Female. Length of thorax:Abdomen::132:172. Seventh abdomi- nal tergite shorter than sixth tergite (82.35:100). Connexival spine of moderate length, arising far beyond the seventh tergite, of moderate length but not reaching the tip of the dorsolateral lobe. Ventrolateral lobe broad and truncate at tip. Distal lobe of seventh abdominal sternite hidden by the broad ventrolateral lobes and directed dorsad. Dorsolateral lobe of seventh abdominal segment moderately long and spine-like, its upper margin from base of con- nexival spine to its tip much less than twice as long as connexival margin of seventh abdominal segment. Incission between the dorso- lateral and ventrolateral lobes of seventh abdominal segment not as deep as broad. In side view ventrolateral lobe of seventh ab- dominal segment longer than the venter of seventh abdominal seg- ment before it. Types: Three females from Hong Kong, China are in the U. S. National Museum. Comparative notes: Every close to, if not identical with, P. harpyia Schmidt from Cambodia which may prove to be a synonym. Distribution: Known only from the type locality. If P. harpyia proves to be a synonym of this species, it will extend its distribution far to the south and westward. 464 Tue UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) tigrina Uhler Fic. 1. Female cotype, dorsal view. Fic. 2. Female cotype, ventral view of sixth and seventh abdominal seg- ments. Note the truncate tip of ventrolateral lobe and that the base of this lobe is broad, its basal margin turned medially, its inner margin nearly meet- ing opposite lobe. Fic. 3. Male of P. harpyia Schmidt from Loas, Tonkin. Fic. 4. Female cotype, ventral view of sixth and seventh abdominal seg- ments. Fic. 5. Male of P. harpyia Schmidt from Laos, Tonkin. Arm of left para- mere seen from rear view. Fic. 6. Male of P. harpyia Schmidt from Laos, Tonkin. Dorsal view of second genital segment. Fic. 7. Male of above, ventral view of genital segments. 465 Genus PrmoMERA AMYOT AND SERVILLE PLATE XVII 466 THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) harpyia Schmidt Plate XVII, figs. 3, 5, 6, 7, and Plate XVIII 1926. Ptilomera harpyia Schmidt, Ent. Mitt. 15(1):65. (Described female from “Cambodja” ). 1927. Ptilomera harpyia, Esaki, Eos, Rev. Esp. Ent. 3(3): 259-260. 1933. Ptilomera harpyia, Hoffmann, Lingnan Sci. Jour. 12 Suppl. 247. 1933. pba: harpyia, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 373, 419-42]. 1941. Ptilomera harpyia, Hoffmann, Lingnan Sci. Jour. 20:34. 1960. Ptilomera harpyia, Matsuda, Univ. Kansas Sci. Bull. 41(2):267. As in other Schmidt types the seventh abdominal segment is mounted on a slide. If this had not been done the type could have been described and illustrated more satisfactorily. We have no specimens. from Cambodia. Fortunately we have found in our Torre-Bueno Collection, “Laos, Tonkin.” The females are very near the cotypes of P. trigrina Uhler from Hong Kong, China, and the type of P. harpyia Schmidt from Cambodia. However, these fe- males from Tonkin are more like the P. harpyia Schmidt type, and we are therefore labeling them P. harpyia Schmidt. Whether the latter is a synonym of P. tigrina or not cannot be determined until we see males that are taken with females of P. tigrina. Size: Schmidt gave the body length of the apterous female type “14.00 mm.” *; greatest width 4.00mm.; length of hind femur 22.00 mm.” We add the following: Width of head 2.05 mm.; width of pronotum 2.10mm.; length of mesonotum 3.43 mm.; greatest width of body across mesoacetabula 3.67 mm. Measurements of an apterous pair from “Laos, Tonkin”: Male. Length 17.14mm.; width of head 2.24mm.; width of pronotum 2.33 mm.; length of mesonotum 3.71 mm.; greatest width of body across mesoacetabula 3.81mm. This we have labeled “Neallotype.” Female. Length 17.71 mm.; width of head 2.10mm.; width of pronotum 2.14 mm.; length of mesonotum 3.62 mm.; greatest width of body across mesoacetabula 3.86 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Female type: Ist:2nd:3rd:4th: :140:32:41:28. Male from Laos: Ist:2nd:3rd:4th::162:40: ?: ?. Female from Laos: Ist:2nd:3rd:4th: :142:32:42:27. * By adding length of the body of the type and slide mount we get 16.66 mm. instead of 14.00 mm. Genus PrmoMERA AMYOT AND SERVILLE 467 Actual Lengths of Leg Segments Ist tarsal Qnd_ tarsal Front leg Femur Tibia segment segment Memalemty pene nia aac 6.95 mm. 6.05 mm. 3.81 mm. 1.52 mm. Male from Laos ........ 8.33 mm. 7.28 mm. Female from Laos ...... 7.86 mm. 6.29 mm. 4.10 mm. 1.57 mm. Middle leg Femur Hind leg Femur Kemalentypey 835 21.52 mm. Female type .......... 22.00 mm. Male from Laos ........ 24.05 mm. Male from Laos ...... 27.95 mm. Female from Laos ...... 20.48 mm. Female from Laos .... 23.67 mm. Female. Relative length of the thorax is to that of the abdomen cannot be determined for the female holotype but for a female from Laos it is 137:172. Connexival spine about as long as seventh ab- dominal tergite and arises at end of seventh connexivum. Both dorsolateral and ventrolateral lobes present. Dorsolateral lobe rather long, more or less spine-like, its upper margin from base of connexival spine to its tip much less than twice as long as con- nexival margin of seventh abdominal segment. Incission between the dorsolateral and ventrolateral lobes not as deep as broad; dorso- lateral lobe broad at base; ventrolateral lobe truncate at tip in types of both P. tigrina Uhler and P. harpyia Schmidt. However, the latter species varies as shown on Plate XVIII figs. 7, 8, 9. Male neallotype. Length of pregenital abdominal segments: Genital segments::102:73. Genital segments: Suranal plate with median lobe small, not surpassing lateral wings which are moder- ately short, their front and rear margins not parallel. Pygofer as seen from below with distal half slender and pointed. Dorsolateral projection of pygofer as seen in dorsal view short and slender. Paramere with shaft longer than arm which turns obliquely laterad as seen from above. In rear view the arm is curved obliquely dorsad, its sides not parallel. Comparatives notes: This species cannot be placed as a synonym of P. tigrina Uhler until the male of the latter is found. Type: The apterous female type bears the labels “Cambodja” and Schmidt’s determination label. They are in the Polish Academy of Sciences, Warsaw, Poland. The apterous male neallotype bears the label “Laos, Tonkin.” Four other males and two females that have been compared with the types of P. tigrina Uhler and P. harpyia Schmidt appear to be the same as the latter. Distribution: In addition to the types we have seen the follow- ing: Laos: “Laos-Tonkin Staudinger and Bank-Haas. 1 female that is like the above. (California Acad. Sci.) “Laos Ban Na Mon. Dec. 12,1919. R. V. de Salvaza” 19 (British Mus. ) 468 ‘THe UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) harpyia Schmidt Fic. 1. Female from Laos, Tonkin, compared with type. Dorsal view. Fic. 2. Side view of sixth and seventh abdominal segments of above female. Fic. 8. Ventral view of sixth and seventh abdominal segments of above fe- male. Fic. 4. Ventral view of seventh abdominal segment of type drawn from slide mount. Fic. 5. Female from Chiengriai, Thailand. Side view of seventh abdomi- nal segment. We have a female like this from Laos. Fic. 6. Female type. Seventh segment of type in dorsal view, drawn from the slide mount. Fic. 7. Female from Trang, Thailand. Side view of seventh abdominal segment. Fic. 8. Another female from above place. Fic. 9. Still another female from Trang, Thailand. Note that only Fig 8 has a truncate end of the ventrolateral lobe as in P. harpyia type. Males do not show such variation. 470 THe UNIVERSITY SCIENCE BULLETIN Thailand: “Trang, Thailand, Namtok Ching. Apr. 18, 1952. Mt. str. M. E. Griffith 2 ¢ ¢,39 9. While the males are undoubtedly P. harpyia (see Plate XVII fig. 3), the females vary, only fig. 8 on plate XVIII is near the typical form. (K. U. Col.). “Chiengmai, Thailand. Doi Sutep. Mt. Stream Mra. 26, 1952, M. E. Griffith” 2°99, 3 nymphs; (K. Us Col.) “Thailand 957d 34 @hienemran Prov. Puang Rapila D. C. Thurman” 29 ¢. (K. U. Col.) Burma: “Myitkyina, Burma Nov. 30, 1945, B. McDermett” 1 ¢ , 2nymphs. (K. U. Col.); “Tingkawk, Burma, May 25, 1944, L. C. Kenitente2 wapeiao ao. Malaya: “Gunong Pulai Johore Feb. 10, 1961, C. H. Fernando” 14, 19; same place’ “Feb, 21, 1961.2 (2 >a) (CGR Colyer “Gunong Pulai, Stream June 1960, C. H. Fernando.” 2¢ 92 (K. U. Col.); “Malay Panin. Selangor” 2 apterous and 1 macropterous & &. (British Museum). Philippines: “C. N. H. Mus. Philippine Zool. Exp. 1946-47. F. G. Werner. leg” 1 g. (Chicago N. H. Mus.). Ptilomera harpyia ceramensis n. subsp. Plate XIX Size: Male holotype (apterous). Length 14.28 mm.; width of head 2.00 mm.; width of pronotum 2.14 mm.; length of mesonotum 2.90 mm.; greatest width of body across mesoacetabula 3.00 mm. Female allotype (apterous): Length 13.90 mm.; width of head 1.95 mm.; width of pronotum 2.00mm.; length of mesonotum 2.86 mm.; greatest width of body across mesoacetabula 3.38 mm. Color: These two specimens have more of the ground color and smaller black areas than is usual and these are not strikingly black but dark brown, and more or less covered by silvery pubescense. The pattern must still be called typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::120:30:38:25. Female: Ist:2nd:3rd:4th::112:26:35:25. Actual Lengths of Leg Segments Ist tarsal 2nd _ tarsal Front leg Femur Tibia segment segment Male eee, cee ee eee 6.57 mm. 5.07 mm, 3.24 mm, 1.33 mm. Hemale sya seater ere 6.19 mm. 5.38 mm, 3.43 mm. 1.38 mm. Middle leg Femur Hind leg Femur Mallet es oe oe ee ee 18.57 mm. Males 2c ian nachenere 21.90 mm. Hemalemmrits une eee 17.14 mm. Hem aleieeen aes ee ene 18.48 mm. Genus PTmoMERA AMYOT AND SERVILLE 471 Male holotype: Length of pregenital abdominal segments: Genital segments::79:78. Genital segments: Suranal plate with median lobe small not surpassing lateral wings which are moderately long and narrow, their front and rear margins not parallel. Pygofer as seen from below long and slender, its tip pointed. Dorsolateral projection slender, barely surpassing the lateral wing of the suranal plate. Shaft of paramere relatively longer and more slender than in P. harpyia harpyia and the arm relatively shorter with its tip turned out instead of inward as viewed from the rear, and its sides not parallel. Female: Length of thorax:Abdomen::115:143. Connexival spine about as long as tergite of seventh abdominal segment and arises from the end of the connexivum. Both dorsolateral and ventro- lateral lobes present. Dorsolateral lobe rather long, its upper mar- gin from base of connexival spine to its tip much less than twice as long as the connexival margin of seventh abdominal segment. Incission between the dorsolateral and ventrolateral lobes not as deep as broad. Ventrolateral lobe not as broad as in typical P. harpyia Schmidt. See Plate XIX fig. 6. Comparative notes: When a longer series of this form are taken from Ceram it may prove to be a distinct species. Types: Holotype male and allotype bear the label “Piroe, Ceram. F. Muir Feb. 1909,” in California Acad. Sci. Mus. Distribution: Known only from the types which came from the Island of Ceram. 472 Fic. Fic. Fic. Fic. ments. Fic. Fic. ments. Lee TS Tue UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) harpyia ceramensis n. subsp. Female allotype, dorsal view. Male holotype, ventral view of genital segments. Male holotype, dorsal view. Female allotype, ventral view of sixth and seventh abdominal seg- Male holotype, rear view of arm of left paramere. Female paratype, side view of sixth and seventh abdominal seg- Genus PrmoMERA AMYOT AND SERVILLE PLATE XIX 473 A474 THE UNIVERSITY SCIENCE BULLETIN Ptilomera sumbaensis n. sp. Plate XX Size: Male (apterous): Length 16.67mm.; width of head 2.57mm.; width of pronotum 2.71mm.; length of mesonotum 3.1mm.; greatest width of body across mesoacetabula 4.52 mm. Female (apterous). Length 14.1mm.; width of head 2.24 mm.; width of pronotum 2.24 mm.; length of mesonotum 2.95 mm.; great- est width of body across mesoacetabula 4.43 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::158:40:48:36. Female: Ist:2nd:3rd:4th: :127:32:38:28. Actual Lengths of Leg Segments 1st tarsal 2nd tarsal Front leg Femur Tibia segment segment Male paratypes) tans 8.81 mm. 7.23 mm. 4.19 mm. 1.81 mm. Female paratype ......... 6.86 mm. 5.76 mm. 3.48 mm. 1.57 mm. Middle leg Femur Hind leg Femur Male paratype .......... 23.57 mm. Male paratype. ar 28.19 mm. Female paratype ........ 18.29 mm. Female paratype ...... 19.81 mm. Male. Length of pregenital abdominal segments:Genital seg- ments: :80:107. Hind coxae reaching rear margin of fifth tergite. Genital segments: Suranal plate with median lobe small, lateral wings plainly surpassing median lobe, front and rear margins of lateral wing nearly parallel. Pygofer, as seen from below rather elongate, its dorsolateral projections short, not extending laterally beyond the lateral wings of suranal plate. Caudal end of pygofer with sides converging and tip roundly pointed and_ obliquely turned dorsad as seen from the side. Dorsolateral projections of pygofer before its middle. Shaft of paramere longer than arm which is turned more or less downward. Female. Length of thorax:Abdomen::115:140. Seventh abdomi- nal tergite nearly as long as the sixth. Connexival spines moderately long, arising from the end of the connexivum. Dorsolateral lobe of seventh segment with its upper margin from base of connexival spine to its tip broadly curved. The dorsolateral lobe broad at base and turned ventrad. The incission between dorsolateral and ventro- lateral lobes deep and narrow. Comparative notes: This species is a near relative of P. pampha- gus and in the key runs out near it for both male and female. The males have the hind coxae reaching to or beyond the rear margin Fic. Fic. Fic. Fic. Fic. Fic. Fic. SUS TON GO LO Genus PTILOMERA AMYOT AND SERVILLE PLATE XX Ptilomera (Ptilomera) sumbaensis n. sp. Female, dorsal view. Male, ventral view of genital segments. Male, dorsal view. Side view of left paramere. Rear view of left paramere arm. Female, ventral view of sixth and seventh abdominal segments. Female, side view of sixth and seventh abdominal segments. 475 476 THe UNIVERSITY SCIENCE BULLETIN of the fifth abdominal tergite which is a distinctive feature. The females of both species have the short dorsolateral lobe directed ventrad and are separated in couplet 42 in the key. Types: Male holotype (apterous), apterous allotype and 42 2, 49 9, apterous paratypes, all bearing the label “Langgai, O. Sumba July 13, 1949. Dr. Buhler and Dr. Sutter.” Holotype, allo- type and 3¢ g¢ and 39@ 9Q paratypes are in the Natural History Museum, Basel, Switzerland. One pair of paratypes are in the K. U. collection. Distribution: Known only from the type locality. Ptilomera maai n. sp. Plate XXI Size: Male holotype (apterous): Length 15.85mm.; width of head 2.24 mm.; width of pronotum 2.33 mm.; length of mesonotum 3.33 mm.; greatest width of body across mesoacetabula 3.71 mm. Female (deilated macropaterous): Length 14.05 mm.; width of head 2.1 mm.; width across humeri 3.05 mm.; greatest width of body across mesoacetabula 3.57 mm. Color: Typical for the genus but these two specimens show more of the ground color than is usual. Structural characteristics: Relative lengths of the antennal seg- ments. Male: Ist:2nd:3rd:4th::150:38:?: ?. Female: Ist:2nd:8rd:4th::125:29:41:31. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Male holotype .......... 8.57 mm. 7.29 mm. 4.52 mm. 1.81 mm. Female allotype ........ 6.66 mm. 5.62 mm. 3.24 mm. 1.43 mm. Middle leg Femur Hind leg Femur Male holotype .......... 24.29 mm. Male holotype ........ 28.81 mm. Female allotype ......... 18.33 mm. Female allotype ....... 20.48 mm Male. Length of pregenital abdominal segments:Genital seg- ments::88:78. Genital segments: Suranal plate with median lobe small, surpassed caudally by the lateral wings, the front and rear margins of which are not parallel. Pygofer as seen from below ap- pearing narrow, dorsolateral projections slender, extending a little laterad of wings of suranal plate. Tip of pygofer nearly pointed. Paramere with exposed shaft shorter than its arm. Hind coxae only reaching rear margin of fourth abdominal tergite. Genus PritoMERA AMYOT AND SERVILLE 477 PLATE XXI Fic. Fig. 3. Fic. Fic. Fic. Fic. ments. Fic. ments. H @ Wir o bo Ptilomera (Ptilomera) maai n. sp. Female allotype (dealated macropterous form). Compare with Male holotype, ventral view of genital segments. Male holotype (apterous), dorsal view. Left paramere. Female allotype, ventral view of sixth and seventh abdominal seg- Female allotype, side view of sixth and seventh abdominal seg- 478 THE UNIVERSITY SCIENCE BULLETIN Female. The length of thorax:Abdomen::115:135. Seventh ab- dominal tergite a little shorter than the sixth. Connexival spines of median length arising from end of connexivum. Dorsolateral lobe moderately long, ventrolateral lobe but faintly indicated. Distal lobe of seventh abdominal sternite of usual size and turned vertically, two short parallel ridges near its base. Comparative notes: In the key to males this species runs out to couplet 21 with P. sumbaensis n. sp. and in our key to females to couplet 50 with P. nunikaensis n. sp. which also occurs in Borneo. Types: The holotype and allotype which are apterous, carry the label “Borneo: Sarawak. Kampong Pueh, Lundu District 690- 1500 m. May 25-31, 1958 T. C. Maa.” These are in Bernice P. Bishop Museum, Honolulu, Hawaii. Distribution: Known only by the above type locality. Ptilomera (Ptilomera) pamphagus Breddin Plate XXII 1901. Ptilomera pamphagus Breddin, Abh. Naturf. Ges. Halle, Bd. 24( 20) :86- 87. Taf. 1 fig. 10. (Described 2, ¢ from Celebes). 1927. Pitlomera pamphagus, Esaki, Eos, Rev. Esp. Ent. 3(3):261. 1933. Ptilomera pamphagus, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372- 373, 518, 420-421, 426-429, figs. 186, 1387. (Gave as synonyms: P. oribasus Breddin, P. laelaps Breddin, P. sumizome Esaki). 1960. Ptilomera pamphagus, Matsuda, Univ. of Kansas Sci. Bull. 41(2):267, 539 (figs. 623, 625) 541 fig. 635. Redescription of the types: Size: Breddin gave the following: Length of female 15.00 mm.; length of middle femur 21.00 mm.; hind femur 22.5 mm.; length of male 16.5mm.; length of middle femur 23.5mm.; hind femur 28.5mm. Our measurements follows: Female. Length 15.23 mm.; width of head 2.26 mm.; width of pronotum 2.43 mm.; length of mesonotum 3.38 mm.; greatest width of body across mesoacetabula 4.12 mm. Male. Length 16.76mm.; width of head 2.43mm.; width of pronotum 2.76 mm.; length of mesonotum 3.33 mm.; greatest width of body across mesoacetabula 4.00 mm. Color: Quite typical for the genus. Head and thorax yellowish brown with the usual black markings. Head with clypeus, antennal sockets and two converging lines on vertex black. The black band on anterior margin of pronotum, anterior lateral margins of meso- and metanotum largely covered by a silvery silky pubescence. The lateral black band of thorax and abdomen with conspicuous band Genus PrmrLOMERA AMYOT AND SERVILLE 479 PLATE XXII Ptilomera (Ptilomera) pamphagus Breddin Fic. 1. Female holotytpe, dorsal view. Fic. 2. Male allotype, ventral view of genital segments. Note that the dorsolateral projections of pygofer extend at least one-fifth of their length be- yond the lateral wings of suranal plate. Fic. 8. Male allotype, dorsal view. Note the long, slender nearly crescent shaped paramere, their shafts and arms are inseparable. Fic. 4. Female type, left side view of sixth and seventh abdominal seg- ments. Fic. 5. Female type, ventral view of sixth and seventh abdominal segments. 480 THe UNIVERSITY SCIENCE BULLETIN of silvery pubescence. Abdominal tergites black, with a median yellowish spot on segments two to seven in female and first six entirely black in male. The black area largely covered by silvery pubescence. Connexivum testaceous. Antennae dark brown to nearly black. Front femur testaceous with two dark longitudinal banks on the dorso-caudal side and a fainter one on the anterior ventral side. Other legs dark brown, the distal ends of femora paler. Entire venter covered with a silvery pubescence. Structural characteristics: Relative lengths of antennal segments. Female: Ist:2nd:3rd:4th::138:33:44.3:30.7. Male: Ist:2nd:3rd:4th: :152:38:46:34. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Remale rn Ge ane eek ete 7.47 mm. 6.29 mm. 4.10 mm. 1.86 mm. Mall Stier ee eterna ee 8.1 mm. 7.14 mm. 4.19 mm. 1.81 mm. Middle leg Femur Hind leg Femur Hlemale sees eee ea oll. Hlemnall eure renee 23.57 mm. INT eal eae ee eS Ste ae 24.29 mm. Maletis eee 29.29 mm. Female type. Length of thorax:Abdomen::130:148. Seventh abdominal tergite about equal to sixth. Connexival spine moder- ately long, arising from end of connexivum. Incission between the dorsolateral and ventrolateral lobes narrow and deep. Dorsolateral lobe short and definitely turned ventrad. Distal lobe of seventh ab- dominal sternite normal in size, directed dorsad and embraced by the lateral lobes. Male allotype. Length of pregenital abdominal segments: Genital segments: :91:83. Genital segments: Suranal plate with median lobe moderate in size, not plainly surpassing lateral lobes, the front and rear margins of which are not parallel. Pygofer as seen from below rather long and slender, its tip bluntly pointed. Dorsolateral projections of pygofer slender and long, in dorsal view extending one-fifth their length beyond the lateral wings of suranal plate. Paramere long and slender, its shaft and arm in one continuous curve, very unlike that of Lundblad’s fig. 137 of a specimen from Sumatra which we have named Ptilomera lundbladi n. sp. Comparative notes: This species is distinct from all others. The long, slender dorsolateral projection of the pygofer separates the male from other species and the lateral profile of seventh abdominal segment separates the female from other species. Types: The types are apterous specimens. The female bears the label “Posso See 900 m. 9, II 95” and Breddin mentioned this as the Genus PrmoMERA AMYOT AND SERVILLE 481 type locality in Celebes for P. pamphagus, and the specimen fits his figure. However, Breddin had labeled this specimen “Ptilomera dorceus Bredd. Typus,” and Lundblad has correctly labeled it Ptilomera pamphagus and added a red label “Typus.” The true type of Ptilomera dorceus, is an apterous female, was stated by Breddin to have come from “Matinang-Kette Siidseite 800-1200 m. (Sar.” Although Dr. Lundblad overlooked this type specimen we found it labeled by him Ptilomera orbasus Breddin, and have re- described and illustrated it. The male type of P. pamphagus bears a label “Posso See” and was recognized by Lundblad as this species and he placed a red “Typus” label and his own determination label upon it. Distribution: Known only from the type localities in Celebes. Ptilomera (Ptilomera) sumizome * Esaki Plate XXIII 1925. Ptilomera sumizome Esaki, Philippine Jour. Sci. 26(1):59, Pl. 1, fig. 7( 24), fig. 8( 2). (Described from South-East Celebes. ) 1927. Ptilomera sumizome Esaki, Eos, Rev. Esp. Ent. 3(3):261. 1933. Ptilomera sumizome Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 419. (Made this a synonym of P. pamphagus Breddin). Dr. Lundblad did not see the types of this species, but placed it as a synonym of P. pamphagus Breddin because it falls within the wide range of variability of the female he ascribed to P. pamphagus Breddin. Since Dr. Esaki described his species as black we ques- tioned this synonymy for we had never seen a truly black species. Thanks to the kindness of Mr. T. Hidaka we have been permitted to study the holotype male and a paratype female of this species which bear the labels “S. O. Celebes T. Elbert 2, 1, 1900, 31” and “Property of the Ent. Lab. Kyushu Univ.” We find the types are not really black, but such a dark chocolate brown that the black spots of the generic pattern are inconspicuous, but as usual, covered with a silvery pile. The structural characters show this to be a distinct species. Redescription of types: Size: Male holotype (apterous): Length 14.29mm.; width of head 2.24 mm.; width of pronotum 2.57 mm.; length of mesonotum 3.1 m.; greatest width of body across mesoacetabula 3.57 mm. Female paratype (apterous): Length 13.14 mm.; width of head 2.05mm.; width of pronotum 2.05mm.; length of mesonotum 2.86 mm.; greatest width of body across mesoacetabula 3.48 mm. * In the Japanese language sumizome means “stained with Chinese ink which is black.” 482 THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) sumizome Esaki Fic. 1. Female paratype. Fic. 2. Male holotype, ventral view of genital segments. Fic. 3. Male holotype, dorsal view. Fic. 4. Male holotype, lateral view of genital segments. Note the elevated caudal section of the dorsum of first genital segment. Fic. 5. Female paratype, ventral view of distal abdominal segments. Fic. 6. Female paratype, lateral view of right side of distal abdominal seg- ments. Fic. 7. Male holotype parameres. GENuS PriLOMERA AMYOT AND SERVILLE PLATE XXIII 484 THe UNIVERSITY SCIENCE BULLETIN Color: Ground color dark chocolate brown, nearly black. The black spot pattern which characterizes this genus very faint but marked by the silvery pubescence which usually cover such spots. Antennae and legs also nearly black, with the longitudinal black lines of the front femur faintly discernible. In lateral view all acetabula black, more or less covered by silvery pubescence. The longitudinal black band of meso- and metapleura covered medially by a band of long silvery pubescence. Ground color of venter of thorax and abdomen black, almost obscured by a frosty pubescence. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:8rd:4th::140:33:42:28. Female: Ist:2nd:3rd:4th::120:31: ?: P. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Male (holotype) ........ 7.43 mm. 6.29 mm. 3.8 mm. 1.62 mm. Female (paratype) ..... 6.29 mm. 5.383 mm. 3.05 mm. 1.43 mm. Middle leg Femur Hind leg Femur Male (holotype) ........ 21.33 mm. Male (holotype) ...... 25.71 mm. Female (paratype) ...... 17.71 mm. Female (paratype ..... 19.28 mm. Male holotype. Length of pregenital abdominal segments:Genital segments: :65:78. Genital segments: Suranal plate with median lobe small, but slightly surpassing the lateral wings; lateral wings short, their anterior and posterior margins not parallel. Pygofer with its dorsolateral projections before its middle and not longer than the lateral wings of suranal plate when viewed from above; distal half of pygofer slender, its caudal tip pointed. Paramere very hairy, its exact shape difficult to see, but both shaft and arm are stout, the latter rounded at tip. (See Plate XXIII, fig. 7.) Distal half of dorsum of first genital segment abnormally elevated. Female paratype. Length of thorax:Abdomen::116:132. Seventh abdominal tergite a little shorter than sixth. Connexival spine of moderate length, arising from the end of the connexivum. Both dorsolateral and caudolateral lobes present, the former short and directed ventrad. An incission separating the lobes moderately narrow. (Plate XXIII fig. 6). Comparative notes: In the key to males this species runs out in couplet 24 with P. harypia Schmidt and in the key to females it runs out in couplet 41 which leads also to P. sumbaensis n. sp. and P. pamphagus Breddin. Types: The types are in the Entomological Laboratory of Kyushu University, Fukuoka, Japan. Genus PrmoMERA AMYOT AND SERVILLE 485 Distribution: Known only from the types which came from South- east Celebes. Ptilomera (Ptilomera) sumatranus n. sp. Plate XXIV Size: Male holotype (apterous): Length 16.29mm.; width of head 2.19 mm.; width of pronotum 2.48 mm.; length of mesonotum 3.71 mm.; greatest width of body across mesoacetabula 4.14 mm. Female allotype (apterous): Length 15.24 mm.; width of head 2.12mm.; width of pronotum 2.24mm.; length of mesonotum 3.48 mm.; greatest width of body across mesoacetabula 4.14 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::168:41:55:32. Female: Ist:2nd:3rd:4th::148:33:41:31. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Mialevan ior curs aes 8.48 mm. 7.33 mm. 4.52 mm. 1.76 mm. Bemale feo ier 7.81 mm. 6.43 mm. 4.33 mm. 1.67 mm. Middle leg Femur Hind leg Femur Male een ear seat ees 25.24 mm. Malevs sce tones an ee 30.10 mm. Hemalerecrnie see toner 0 21.19 mm. Hemale i) ao ee 23.10 mm. Male. Length of pregenital segments of abdomen:Genital seg- ments::84:72. Genital segments: Suranal plate with median lobe of moderate size. Lateral wings of suranal plate not surpassing caudally the median lobe, their front and rear margins not parallel. Pygofer, as seen from below, normal in shape and its caudal tip blunt. Dorsolateral projections of pygofer slender, pointed and ex- tending laterally about as far as lateral wings of suranal plate; shaft of paramere longer than the laterally turned arm. In rear view the arm has a distinct keel on its lower margin as shown on Plate XXIV fig. 4. Female. Length of thorax:Abdomen::130:158. Seventh abdomi- nal tergite subequal to the sixth tergite. Connexival spine of moderate length arising from end of connexivum. Dorsolateral lobe short, directed obliquely downward, separated from ventro- lateral lobe by a narrow deep incission. The distal lobe of seventh abdominal sternite of usual size and shape and partly covered by the lateral lobes. If seen from the rear its distal margin is truncate and not pointed. 486 Fic. Fic. Fic. Fic. Fic. Fic. ments. Fic. ments. oo to THe UNIVERSITY SCIENCE BULLETIN Ptilomera ( Ptilomera) sumatranus n. sp. Female allotype, dorsal view. Male holotype, ventral view of genital segments. Male holotype, dorsal view. Rear view of left paramere. Side view of right paramere. Female allotype, ventral view of sixth and seventh abdominal seg- Female allotpe, left side view of sixth and seventh abdominal seg- GeNus PriLoMERA AMYOT AND SERVILLE PLATE XXIV 487 488 THE UNIVERSITY SCIENCE BULLETIN Comparative notes: This is one of several closely related species from Sumatra that would be puzzling if not accompanied by males. Types: An apterous holotype male and an apterous female allo- type and two paratypes ( ¢- @ ) bear the label: “Sud-Sumatra, Lampongs. Mt. Tenggamoes. M. E. Walsh, 15-2000 ft.” The holotype and allotype are in the Natural History Museum, Basel, Switzerland. Doctors Sutter and Keiser have given the paratypes to University of Kansas collection. Distribution: Known only from the type locality. Ptilomera (Ptilomera) lundbladi n. sp. Plate XXV Lundblad’s fig. 136B and 137 were reproduced from Archiv fiir Hydrobiologie, Suppl. Bd. 12:427-428 (1933). In fig. 186B is a side view figure of the seventh segment of a female that came from Subang Ajam, Sumatra that Dr. Lundblad figured as only a varia- tion of P. pamphagus Breddin. In fig. 137 he figured the genitalia of a male from the same place and labeled it P. pamphagus. Had he examined the paramere of the type of P. pamphagus he would have found that the male from Celebes is quite unlike his male from Sumatra. Therefore, we offer the name P. lundbladi for the species indicated above identified by Lundblad as Ptilomera pam- phagus which it cannot be. By the generosity of Dr. O. Lundblad of Stockholm, Sweden, we have been permitted to reproduce his text figures 136 and 137 on this plate. Since figure 136B and figure 137 are not P. pamphagus Breddin but an undescribed species we have named it in his honor. Ptilomera (Ptilomera) laelaps Breddin Plate XXVI 1901. Piéilomera laelaps Breddin, Abh. Naturf. Ges, Halle, 24:20, 88-89. Tat. 1, fig. 13 (Lists on p. 20, described male from Celebes on pp. 88-89). 1903. Ptilomera laelaps, Breddin, Societas Entomologica 17(19):147. (De- scribed female from Samanga, S. Celebes). 1927. Ptilomera laelaps, Esaki, Eos, Rev. Esp. Ent. 3(3) :260. 1933. Ptilomera laelaps, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372-73, 419-21. (Gave as synonym of P. pamphagus). In his paper “Die Hemipteren von Celebes” Dr. Breddin de- scribed the male from “S. Loka, 1000-3000m. Okt. 95 (Sar); Makassar (Coll. Breddin).” In 1903 he described the female from Genus PrmoMERA AMYOT AND SERVILLE 489 PLATE XXV Fig. 137. Ptilomera pamphagus Bredd, g (apter). A Kniegelenk des Vorderbeins; B Genitalkapsel von oben (Penis schwarz; linker Genitalgriffel entfernt); CC Distalende der Genitalkapsel; D linker Genitalgriffel; E Analplatte. Nach einem Expl. aus Subang Ajam, Tjurup, Sumatra. a, Fig. 136. Ptilomera pamphagus Bredd. 99 (apter). A die Type von pamphagus; B nach einem Expl. aus Subang Ajam, Sumatra; C—D aus dem Wai Negri, Sumatra; E aus Balige, Sumatra; F aus Umbilin, Sumatra; G die Type von oribasus; H aus Balige, Sumatra; I die Type von laelaps. 490 Fic. spines. Fic. ments. Fic. Fic. Fic. Fic. Fic. ments. aw SDA E>. Go THe UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) laelaps Breddin Female allotype, dorsal view. Note the very short connexival Female allotype, ventral view of sixth and seventh abdominal seg- Male holotype, dorsal view. Male holotype, ventral view of genital segments. Male holotype, dorsal view of genital segments. Male holotype, rear view of arms of the parameres. Female allotype, side view of sixth and seventh abdominal seg- Genus PrmomMERA AMYOT AND SERVILLE PLATE XXVI 49] 492 THE UNIVERSITY SCIENCE BULLETIN S. Celebes (Samanga). Fortunately, we found in our Kirkaldy Col- lection 1 apterous male and 3 apterous females and one macropter- ous male bearing the identical label of the female type “Samanga S. Celebes Nov. 1895 H. Fruhstorfer” and one apterous male bear- ing the label “S. Celebes Bua-Kraeng 5000 ft. Feb. 1896, H. Fruhstor- fer.” The males of our series are like the male type and the three females are like the female type and show no recognizable varia- tion in the length of the connexival spine. Redescription Size: Female type (apterous): Length 14.38 mm.; width of head 2.02mm.; width of pronotum 2.10mm.; length of mesonotum 2.90 mm.; greatest width of body across mesoacetabula 3.62 mm. Male (apterous): Bearing the same collector’s label as the female type. Length 15.24mm.; width of head 2.10mm.; width of pro- notum 2.43 mm.; length of mesonotum 3.19 mm.; greatest width of body across mesoacetabula 3.71 mm. Male (macropterous): From the same place as above. Length including wings 15.14mm.; width across head 2.10mm.; width across humeri 3.14 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Female type (apterous): Ist:2nd:3rd:4th::125:28:40:27. Male (apterous): Ist:2nd:3rd:4th: :132:33.5:43:30. Male (macropterous): Ist:2nd:3rd:4th: :123:32:40:29. Actual Lengths of Leg Segments Ist tarsal 2nd _ tarsal Front leg _ Femur Tibia segment segment Hemalestypei seen 6.90 mm. 5.71 mm. 3.33 mm. 1.48 mm. Male (apterous) ........ 7.29 mm. 6.19 mm. 3.43 mm. 1,52 mm. Male (macropterous) .... 6.86 mm. 5.81 mm. 3.10 mm. 1.43 mm. Middle leg Femur Hind leg Femur Female type). 2% oe 19.1 mm. Female type .......... 20.5 mm. Male (apterous) ........ 20.47 mm. Male (apterous) ...... 24.00 mm. Male (macropterous) .... 18.57 mm. Male (macropterous) .. 21.67 mm. Male. Length of pregenital abdominal segments:Genital seg- ments: :80:62. Genital segments: Suranal plate not large but sur- passing caudally broad short lateral wings which have their front and rear margins not parallel. Pygofer with its dorsolateral pro- jections short; pygofer, as seen from below, not conspicuously broad beyond its middle, its venter of usual shape and its tip roundly pointed. Paramere with shaft relatively slender and not sharply separated from its arm, not shaped like that of P. harpalos Genus PrmomMERA AMYOT AND SERVILLE 493 Schmidt which runs out also in couplet 19 of the key. First genital segment with its dorsal line but little elevated in caudal half as seen from side; venter short, its median longitudinal kee] broad. Female. Length of thorax: Abdomen::128:148. Seventh abdomi- nal tergite as long as sixth. Connexival spine stout, short, plainly shorter than seventh tergite, not half as long as upper margin of dorsolateral lobe from base of connexival spine to its tip. Dorso- lateral and ventrolateral lobes short, the latter represented by a convex curve in caudal margin and separated from the former by a broad shallow incission. Dorsolateral lobe directed obliquely downward. Distal lobe of seventh sternite moderately small, its sides slightly converging and slightly concave, its distal end slightly convex. Comparative notes: This species is related to P. harpalos Schmidt and runs out to couplet 19 in the key to males and to couplet 15 in the key to females. Types: One apterous male one apterous female. Types are in the Deutsches Entomologisches Institut, Berlin. Distribution: Besides the types we have one apterous male and three apterous females and one macropterous male bearing the identical label of the female type. “Samanga, S. Celebes Nov. 1895 H. Fruhstorfer” and one apterous male bearing the label “S. Celebes, Bua-Kraeng 5000 ft. Feb. 1896 H. Fruhstorfer.” Ptilomera (Ptilomera) harpalos Schmidt Plate XXVII 1926. Ptilomera harpalos Schmidt, Ent. Mitt. 15(1):66. (Described from Sumatra ). 1927. Ptilomera harpalos, Esaki, Eos, Rev. Esp. Ent. 3(3) :260. 1933. Ptilomera harpalos, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 373, 419, 420. 1960. ea harpalos, Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 539, : Redescription: Size: Schmidt gives the length of male type 15.5 mm.; greatest width 4.00 mm.; hind femur 27.00mm. Length of female 15.5 mm.; greatest width 3.5-4.0mm. (he had two females); hind femur 23.00 mm. We give the following measurements for the types. We must accept Schmidt’s measurement for the length of body because he removed and mounted the genital segments on slides. Male type (apterous): Length 15.5 mm.; width of head 2.14 mm.; width of pronotum 2.33 mm.; length of mesonotum 3.67 mm.; great- 19—1367 494 THE UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) harpalos Schmidt Fic. 1. Female from Sumatra, Si Rambé. Dorsal view. Fic. 2. Male type. Genital segment drawn from slide mount. Fic. 3. Male from Sumatra, Si Rambé. Dorsal view. Fic. 4. Female from Sumatra, Si Rambé. Ventral view of sixth and seventh abdominal segments. Fic. 5. Male from Sumatra, Si Rambé. Enlarged drawing of paramere of figure 8. Fic. 6. Female from Sumatra, Si Rambé. Side view of sixth and seventh abdominal segments. Fic. 7. Male from Sumatra, Si Rambé. Ventral view of genital segments. Fic. 8. Female type. Dorsal view of abdominal tip drawn from slide mount. Genus PrmoMERA AMYOT AND SERVILLE 495, PLATE XXVII 496 THE UNIVERSITY SCIENCE BULLETIN est width of body across mesoacetabula 3.90mm. Female type (apterous). Length 15.5mm.; width of head 2.05 mm.; width of pronotum 2.05 mm.; length of mesonotum 3.43 mm.; greatest width of body across mesoacetabula 3.86 mm. We give below the meas- urements of a pair from our Kirkaldy Collection bearing the label “Dr. Hagen, Toba-meer Sumatra inter.” Male (apterous). Length 14.29 mm.; width of head 2.1 mm.; width of pronotum 2.31 mm.; length of mesonotum 3.48 mm.; greatest width of body across mesoacetabula 3.57 mm. Female (apterous). Length 15.33 mm.; width of head 2.00mm.; width of pronotum 2.12mm.; length of mesonotum 3.38 mm.; greatest width of body across mesoacetabula 3.62 mm. Color: Typical of the genus. Structural characteristics: Relative lengths of antennal segments. Male type: Ist:2nd:3rd:4th::147:35: P: P. Female type: Ist:2nd:3rd:4th: :138:31:45:32. Male (Kirk. Coll.): Ist:2nd:3rd:4th: :137:37:50:30. Female (Kirk. Coll.): 1st:2nd:3rd:4th: :120: 28:37:29. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Maleatype eerste n tice 8.00 mm. 7.14 mm. 4.19 mm. 1.67 mm. Male (Kirk. Goll.) .... ©. 7.60 mm. 6.60 mm. 3.90 mm. 1.60 mm. Female: type 2.7.4 4: 7.62 mm. 6.57 mm. 4.00 mm. 1.62 mm. Female (Kirk. Coll.) .... 6.80 mm. 5.70 mm. 3.70 mm. 1.52 mm. Middle leg Femur Hind leg Femur Maletbypeis we ao a oe 22.38 mm. Male type: a2 ee 28.1 mm. Malex@Kink Coll) ae 19.38 mm. Male (Kirk. Coll.) ...:. 22.86 mm. Female type ............ 21.33 mm. Female type .......... 23.09 mm. Female (Kirk. Coll.) ..... 17.62 mm. Female (Kirk. Coll.) ... 19.52 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::83:64. Genital segments: Suranal plate not large but sur- passing caudally short lateral wings which have their front and rear margins not parallel. Pygofer with its dorsolateral projections short; pygofer as seen from below not conspicuously broad beyond its middle, its venter of usual shape, its tip roundly pointed. Para- mere with exposed part of shaft stout, short and broad, its lateral margin abruptly turned mesally to form the base of the arm which turns laterally. First genital segment with its dorsal line nearly straight, caudal half only elevated, venter short, its median longi- tudinal keel broad. Female. Length of thorax:Abdomen::125:160. Seventh abdomi- nal tergite about as long as sixth. Connexival spine relatively short, slender and sharp pointed, usually at least half as long as upper Genus PrmomMERA AMYOT AND SERVILLE 497 margin of dorsolateral lobe from base of connexival spine to its tip. Dorsolateral lobe stout and short, directed obliquely downward; ventrolateral lobe short, nearly rectangular at tip and separated from dorsolateral lobe by a broad shallow incission that is deeper than that of P. laelaps Breddin. Comparative notes: This species is a near relative of P. laelaps Breddin from which it differs in both sexes. The paramere has the exposed portion of its shaft short and broad, with its distolateral margin abruptly turned mesad to form the base of arm. The para- mere in P. laelaps is different as seen in the key. The female has a longer and more slender connexival spine than in P. laelaps. Types: One male and one female types bear the label “Dohrn Sumatra. Sinabong.” One female with the label “Soekaranda, Jan. 1894 Dohrn.” All apterous. All types are at the Polish Academy of Sciences, Warsaw, Poland. Distribution: Besides the types we have the following from Sumatra. “Dr. B. Hagen Toba-meer, Sumatra inter” 13,29 9, (all apterous) (K. U. Kirkaldy Collection); “Sumatra Si-Rambe XII 90 III 91 E. Modigliani’ 1 g,39¢ ¢, (apterous) (K. U. Kirkaldy Collection ); “Sibolga Tapianoeli Sumatra 9, 16, 31 v. d. Meer Mohr” 29 9 (apterous) (K. U. Coll.); Near Sembahe, E. coast Sumatra 6.7 31 v.d. Meer Mohr, 1 ¢ (apterous) (K. U. Coll.). Ptilomera (Ptilomera) oribasus Breddin Plate XXVIII, figs. 1, 2, 3, 5, 6, 7 1901. Ptilomera oribasus Breddin, Abh. Naturf. Ges. Halle, Bd. 24:20,88. Taf. 1 fig. 122. (Described from Celebes). 1927. Ptilomera oribasus, Esaki, Eos, Rev. Esp. Ent. 3(3):260. 1933. Ptilomera oribasus, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 373, 419-21. (Gave as synonym of P. pamphagus Breddin). Breddin’s collection contains two female specimens, each with a red “type” label. One bears the written label “Celebes. Sud. Vorberge der Takalekadjo-Kette 1000 m. 7 II 95 Leg. Sarasin,” and “Ptilomera oribasus Breddin Typus,” in Breddin’s handwriting. This specimen is the one Breddin drew and also the one Lundblad drew. The other female specimen bears the labels, “Celebes Natinang- Kette Seite 800-1200 m. (Sar.).” It also bears a label “Ptilomera oribasus Bred.” which is not Breddin’s writing. It looks like Dr. Lundblad’s printing. He is the only one who has written on these types. However, this female fits Breddin’s figure of Ptilomera dorceus and bears the type locality label of this species. Dr. Lund- blad must have overlooked this specimen because he said on page 498 Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. I Ct THe UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) oribasus Breddin Female allotype, dorsal view. Male neallotype, dorsal view. Male neallotype, ventral view of genital segments. (Brit. Mus.) Male neallotype, dorsal view of left paramere. Female type, side view of sixth and seventh abdominal segments. Female type, ventral view of sixth and seventh abdominal segments. Ptilomera (Ptilomera) dorceus Breddin Female type, dorsal view. Female type, lateral view of sixth and seventh abdominal segments. Female type, ventral view of sixth and seventh abdominal segments. 499 Genus PritoMEeRA AMYOT AND SERVILLE PLATE XXVIII 500 THe UNIVERSITY SCIENCE BULLETIN 419 of his great work “Zur Kenntnis der Aquatilen und Semi- aquatilen Hemipteren von Sumatra, Java, und Bali 1933,” that the only species of Ptilomera unknown to him were P. tigrina, P. dorceus, and P. sumizome. We have now before us the types of all three of these as well as the type of P. oribasus Breddin which is a female. In addition we have had the good fortune of finding the unde- termined material of the British Museum a male and female of this species and have labeled the female “P. oribasus Breddin” and “compared with Type.” We have labeled the male “P. oribasus” “Neallotype.” Both specimens bear the label “F. C. Drescher Zuid Celebes Nanggala 900m. Rantepao VI 1937.” Therefore, a more complete description of the species follows: Size: Female type (apterous): Breddin says length is 14.00 mm.; middle femur 18.5 mm.; hind femur 21.5mm. Our measurements for this type: Length 13.81 mm.; width of head 2.05 mm.; width of pronotum 2.17 mm.; length of mesonotum 3.05 mm.; greatest width of body across mesoacetabula 3.90 mm. Male neallotype (apterous): Length 15.71 mm.; width of head 2.31 mm.; width of pronotum 2.43mm.; length of mesonotum 3.29 mm.; greatest width of body across mesoacetabula 4.09 mm. Apterous female taken with neallotype. Length 14.86 mm.; width of head 2.14mm.; width of pronotum 2.24mm.; length of meso- notum 3.19 mm.; greatest width of body across mesotabula 4.19 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Female type: Ist:2nd:3rd:4th::127:27: ?: ?. Male neallotype: Ist:2nd:3rd:4th::152:35:48: ?. Female came with neallotype lost antennae. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Hemalextypessss 4 ieee: 7.14 mm. 6.1 mm. 3.57 mm. 1.52 mm. Male neallotype ........ 7.85 mm. 6.66 mm. 4.38 mm. 1.67 mm. Female came with above, 7.00 mm. 5.85 mm. Middle leg Femur Hind leg Femur Hemalestypeni ce. far 18.81 mm. Female type .......... 22.14 mm. Male neallotype ......... 227 lem: Male neallotype ....... 28.57 mm. Female came with Female came with nealloty pee) cya oe 18.81 mm. neallotype: 4. ons 20.81 mm. Female type (apterous). Length of thorax: Abdomen: :120:137. Female came with male neallotype (apterous). Length of thorax: Abdomen: :120:140. GreNus PrmoMERA AMYOT AND SERVILLE 501 Female (apterous). Length of thorax:Length of abdomen: :122: 146. Seventh abdominal tergite as long as sixth. Connexival spine arising from end of seventh connexivum and at least half as long — as upper margin of dorsolateral lobe from base of connexival spine to its tip. An incission separating the dorsolateral and ventro- lateral lobes of seventh abdominal segment wide. Dorsolateral lobe short, slender and sharply pointed. Distal lobe of the seventh abdominal sternite of usual size and embraced by lateral lobes. Male neallotype (apterous). Length of pregenital abdominal segments:Genital segments::75:73. Genital segment: Suranal plate with median lobe not large, but surpassing lateral wings caudally; lateral wing with front and rear margins nearly parallel, the rear margin a little longer than width of its tip. Pygofer with its dorso- lateral projections short but surpassing lateral wings of suranal plate and ending in a dense brush of long hair. Pygofer, as seen from below, not broad beyond its middle, but appearing rather elongate and its tip pointed. Venter of genital segments without a complete keel on the first genital. Comparative notes: This species which came from Celebes ap- pears to have its nearest relatives in New Guinea. See couplet 17 of key to males. However, in the key to females, which is certainly not a phylogenetic key, it is not far removed from two New Guinea forms before it and two New Guinea species after it. Types: The apterous female type bears the label “Celebes Sudl. Vorberge der Takalekadjo-Kette 1000 m. 7 II 95 Leg. Sarasin.” They are in the Deutsches Entomologisches Institut, Berlin. The apterous male neallotype and a female taken with it bear the label “F. C. Drescher. Zuid Celebes Nanggala 900 m. Rantepao VI 1937. These are in the British Museum. Distribution: Known only from the above type localities. Ptilomera (Ptilomera) dorceus Breddin Plate XXVIII, figs. 4, 8, and 9 1901. Ptilomera dorceus Breddin, Abh. Naturf. Ges. Halle, 24:20, 87-88. Taf. lla, 1lb. (Described 2 from Celebes and figured genital segments. ) 1927. Ptilomera dorceus, Esaki, Eos, Rev. Esp. Ent. 3(3):260. 1933. Ptilomera dorceus, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:372, 420. Dr. Lundblad did not know P. dorceus Breddin but believed it to be the same as P. dromas Breddin 1903 which he knew. How- ever, since we found what we believe to be four good species on his fig. 136 (all nine drawings of which he thought were variations of P. pamphagus Breddin), we are inclined to question some of the 502 THe UNIVERSITY SCIENCE BULLETIN variations of P. dromas Breddin in Lundblad’s fig. 134. We believe that since we have keyed out the female type of P. dorceus Breddin, we must consider it a species until it is taken together with some males from Celebes that prove otherwise. We therefore offer the following notes on this female type: Size: Breddin gave its body length as 14.00 mm., middle femur 17.5mm., hind femur 19.5mm. Our measurements: Length of body 14.29mm.; width of head 2.05mm.; width of pronotum 2.12mm.; length of mesonotum 2.90 mm.; greatest width of body across mesoacetabula 3.67 mm. Color: Typical for the genus. This specimen has two broad black converging bands on vertex. The black bands on anterior margin of pronotum medially joined. The black band on the side of meso- thorax separated by a longitudinal silvery dense pubescence, be- neath which there shows no yellow stripe, the usual black band is almost obliterated by white pubescence that continues on the venter. Structural characteristics: Relative length of antennal segments. Female type: Ist:2nd:3rd:4th: :123:29:41:27. Actual Lengths of Leg Segments Ist tarsal 2nd _ tarsal Front leg Femur Tibia segment segment Female type ........ .. 6.86 mm. 5.9 mm. 3.29 mm. 1.48 mm. Middle leg Femur Hind leg Femur Female type ............ 18.19 mm. Female type .......... 19.52 mm. Female. Length of thorax: Abdomen::112:152. Seventh abdomi- nal tergite slightly longer than sixth. Connexival spine moderately long, arising from end of connexivum. Incission between dorso- and ventrolateral lobes not as deep as broad. Dorsolateral lobe of seventh abdominal segment only moderately long, its upper margin from base of connexival spine to its tip much less than twice as long as connexival margin of seventh abdominal segment. The dorso- lateral lobe rather slender and spine-like. The ventrolateral lobe short, bluntly pointed; in side view it is shorter than the venter of the seventh abdominal segment; in ventral view the ventrolateral lobe is not broad. Distal lobe of seventh abdominal sternite of usual size and shape, directed vertically and embraced by the lateral lobes. We regret that the male of this Celebes species is unknown. Comparative notes: This species must be a near relative of P. dromas Breddin from Java but until we have both males and females of this species from Celebes we cannot determine how near. Genus PrmoMERA AMYOT AND SERVILLE 503 Type: Apterous female from “N. Celebes, Matinang-Kette Sud- seite 800-1200 m. (Sar).” In the Deutsches Entomologisches Institut, Berlin. Distribution: Known only from the above place. Ptilomera (Ptilomera) dromas Breddin Plates XXIX and XXX 1901. Ptilomera dromas Breddin, Abh. Naturf. Ges. Halle, Bd. 24, Taf. 1, 14 a, b. (While his title was “Die Hemipteren von Celebes” P. dromas n. sp. came from Java as he explained in 1903). 1903. Ptilomera dromas Breddin, Soc. Ent. 17(9):147. (Refers to above and described male and female from East and South Java). 1905. Ptilomera dromas Breddin, Mitt. Naturh. Mus. Hamburg, 22:131-32. (Redescribed both sexes from Java. ). 1915. Ptilomera dromas, Bergroth, Zool. Med. Leiden, 1-3:123. (Noted that Distant and old authors figured male as female and vice versa in this genus ). 1927. Ptilomera dromas, Esaki, Eos, Rev. Esp. Ent. 3(3):260. 1929. Ptilomera dromas, Dover, Treubia 10(1):69. 1933. Ptilomera dromas, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:423, Taf. 13, figs. 134-35. (Suggested that P. argus Breddin and P. asbolus Breddin are synonym and we agree ). 1960. Ptilomera dromas, Matsuda, Univ. Kansas Sci. Bull. 41(2):267, 537, figs. 539, 610, 622. The following synonymies also belong here. 1903. Ptilomera argus Breddin, Societas Entomologica 17(19):147-48. (De- scribed male from W. Java). 1905. Ptilomera argus Breddin, Mitt. Naturh, Mus. Hamburg. 22:134. (Re- peated description of male in 1903). 1927. Ptilomera argus, Esaki, Eos Rev. Esp. Ent. 3(3):260. 1933. Ptilomera argus, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:371. (Saw type and thought it is only an unusually small example of P. dromas Breddin ). 1905. Ptilomera asbolus Breddin, Mitt. Naturh. Mus. Hamburg 22:132-34, figs. 10-11. (Drawings of male and female genital segments of speci- mens from Java). 1927. Ptilomera asbolus, Esaki, Eos, Rev. Esp. Ent. 3(3):260. 1933. Ptilomera asbolus, Lundblad, Arch. Hydrobiol. Suppl. Bd. 12:373, 421. (Said is syn. of P. dromas Breddin). When we first examined the male and female types of P. dromas Breddin from East Java, a male of P. argus Breddin from West Java, and a female of P. asbolus Breddin from Java we found that while P. asbolus was described and figured for both sexes the male is not now present in the Breddin collection. We also thought that P. dromas and P. argus were different species and that P. asbolus was a synonym of P. argus. However, when we tried to key out P. dromas and P. argus we found it impossible to do so and have reached the conclusion that both P. argus and P. asbolus are synonyms of P. dromas. We offer the following descriptions of various types. 504 THE UNIVERSITY SCIENCE BULLETIN PLATE XXIX Ptilomera (Ptilomera) dromas Breddin Fic. 1. Female type, dorsal view. Fic. 2. Female type, ventral view of sixth and seventh abdominal seg- ments. Fic. 3. Male type, dorsal view of genital segments. Fic. 4. Male type, dorsal view. Fic. 6. Male type, arm of left paramere as seen from the rear. Fic. 7. Male type, genital segments in ventral view. Fic. 8. Female type, sixth and seventh abdominal segments as seen from left side. Ptilomera (Ptilomera) asbolus Breddin Fic. 5. Female type, showing sixth and seventh abdominal segments from left side. Ptilomera (Ptilomera) argus Breddin Fic. 9. A female from Buitenzorg, Java, that was taken with males that are the same as the P. argus type. Drawing shows sixth and seventh abdomi- nal segments seen from left side. Fic. 10. The same female showing a ventral view of the same segments. Genus PrmomMERA AMYOT AND SERVILLE PLATE XXIX 505 506 THe UNIVERSITY SCIENCE BULLETIN PLATE XXX Ptilomera (Ptilomera) dromas Breddin Showing that Ptilomera argus Breddin is a synonym and that the species Fic. Fic. Fic. Fic. Fic. Fic. Fic. DW wo bo ls: occurs in Borneo as well as in Java. Female from West Borneo, dorsal view. Another female from the same place. Male type of P. argus Breddin. Female from Buitenzorg, Java. Male type of P. argus Breddin, dorsal view of genital segments. Arm of left paramere of above seen in rear view. Female from West Borneo showing venter of sixth and seventh ab- dominal segments. Fic. 8. Left side view of above. Fic. 9. Male type of P. argus Breddin showing ventral view of genital seg- ments. 507 GENus PritoMERA AMYOT AND SERVILLE PLATE XXX 508 THe UNIVERSITY SCIENCE BULLETIN Types of Ptilomera dromas Breddin Size: Male type (apterous): Length 16.67 mm.; width of head 2.29mm.; width of pronotum 2.67mm.; length of mesonotum 3.71 mm.; greatest width of body across mesoacetabula 4.20 mm. Length of pregenital abdominal segments:Genital segments: :93:84. Female type (apterous): Length 15.24 mm.; width of head 2.10mm.; width of pronotum 2.21mm.; length of mesonotum 3.10 mm.; greatest width of body across mesoacetabula 4.00 mm. Length of thorax: Abdomen:123:161. Color: Color and color pattern are typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male: Ist:2nd:3rd:4th::158:37: P: ?. Female: Ist:2nd:3rd:4th::138:30: P: ?. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Malestype eas oe one: 8.33 mm. 7.14 mm. 4.43 mm. 1.76 mm. Hemale type) 2). 5-4-2 7.24 mm. 6.33 mm. 3.71 mm. 1.57 mm. Middle leg Femur Hind leg Femur Malesty pest bate 22.86 mm. Maleitype les 28.10 mm. Femalevity penn ee ee aes tes 19.19 mm. Femalettype) = 2). canes 21.19 mm. The male type of P. dromas has the seventh abdominal tergite plainly shorter than the three preceding segments together. Length of pregenital abdominal segments:Genital segments: :93:84. Female. Length of thorax: Abdomen::123:161. Seventh abdomi- nal tergite as long as sixth. Male type of Ptilomera argus Breddin (apterous ) Size: Length 14.05 mm.; width of head 1.95 mm.; width of pro- notum 2.05 mm.; length of mesonotum 2.90 mm.; greatest width of body across mesoacetabula 3.14 mm. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments. Male type: Ist:2nd:3rd:4th::120: P: ?. The type has now only the first segment. Actual Lengths of Leg Segments 1st tarsal 2nd tarsal Front leg Femur Tibia segment segment Brontllecya ee 6.75 mm. 5.60 mm. 2.95 mm. 1.30 mm. Middle wleg! => 4.4 se 17.38 mm. Pindeleg.a aver sae eee 20.24 mm. Genus PrmoMERA AMYOT AND SERVILLE 509 The male type of P. argus has the seventh abdominal tergite nearly as long as the three preceding tergites together. We thought this character might separate P. dromas and P. argus but it proved to be variable. Length of pregenital abdominal segments:Genital segments: :76:64. Female type of Ptilomera asbolus Breddin (apterous ) Size: Length 17.19 mm.; width of head 2.05 mm.; width of pro- notum 2.10 mm.; length of mesonotum 3.14 mm.; greatest width of body across mesoacetabula 3.71 mm. Length of thorax:Abdomen:: 117:145. Color: Typical for the genus. Structural characteristics: Relative lengths of antennal segments: Ist-2ndiord: 4th 1272330)? : P. Actual Lengths of Leg Segments 1st tarsal 2nd _ tarsal Front leg Femur Tibia segment segment Mrontelege ci ee eee ye 6.90 mm. 5.71 mm. 4.10 mm. 1.52 mm Mid dlegle genes ane: sae 17.62 mm. indeles a tee si rene 20.00 mm. Since none of the types has an entire antenna we give the com- parative lengths of antennal segments of specimens in our collec- tion that most closely resemble the types. The one that resembles the type of P. dromas: male |st:2nd:3rd: Ath::170:37:53:35; female 140:31:44:381. The one that resembles the type of P. argus: male 1st:2nd:3rd: 4th: :135:33:43:32; female 124:27:39:30. The above specimens came from the same place “Buitenzorg, Java.” All three of these types have the following characteristics in common: Male genital segments: Suranal plate with median lobe small, its lateral wings slightly surpassing median lobe; front and rear margins of lateral wings not parallel. Pygofer with its dorsolateral projections as seen from above short, apex of pygofer blunt. As seen from below pygofer not conspicuously broad beyond its mid- dle; paramere with both shaft and arm rather stout. Female. Length of thorax:Abdomen::118:149. Seventh tergite is nearly equal to sixth tergite in length. Connexival spines moderately long, arising from end of connexivum; dorsolateral lobe moderately long and spine-like, its upper margin from base of con- nexival spine to its tip much less than twice as long as upper con- nexival margin. With a ventrolateral lobe that is triangular; incis- 20—1367 510 ‘THE UNIVERSITY SCIENCE BULLETIN sion between dorsal and ventrolateral lobes broad and shallow; distal lobe of sternite of usual size and usually embraced by lateral lobes. Comparative notes: This is the only species that occurs in Java. It was this slightly variable species and P. pamphagus Breddin complex of several species that made Dr. Lundblad believe that the taxonomy of this genus was quite hopeless. Types: The types of P. dromas, P. argus and P. asbolus are all in the Breddin Collection of the Deutsches Entomologisches Institut in Berlin, Germany. Distribution: Male and female types of P. dromas bear the label “Ost-Java Tengger Gebirg, H. Fruhstorfer.”. The male type of P. argus bears the label “Java occident. Sukabumi 2000° 1893, H. Fruhstorfer.” The female type of P. asbolus bears the label “Java Foerster’; Breddin added Buitenzorg (Kraepelin). Beside the types we have seen the following apterous specimens: Java: “Buitenzorg-Java, Apr. Dec. 96, D. G. Fairchild” 22 3, 40 9 (K. U. Torre-Bueno Coll.);2 3 ¢,292 9 bearing the same label we have marked “P. argus compared with type’; “Muller, Java’ 3g g¢,1¢ (K. U. Kirkaldy Coll.); “Goeneg Niseere, Dam- pang W. Java 3, 1926 Mrs. M. E. Walsh” 1g¢,3¢9 92 (K. U.); “Java Fruhstorfer G. Severin” 19 (we labeled this P. asbolus, compared with type); “Giava, Sindana Java” 19 (K. U. Coll.); “Buitenzorg Java’ 39 9 (K. U. Kirkaldy Coll.); “Java Dr. H. Bos” 1¢@ (K. U. Kirkaldy Coll.); “Reinn Java” 19 (K. U. Kirkaldy Coll.); “C. N. Java Moeria Mts. 1939, Tjolo 20-24 X, M. A. Lief- tinck” 1 ¢,19 (K. U. Coll.); “Buitenzorg Java 3. 09 Bryant and Palmer Coll.” 1g (K. U. Torre-Bueno Coll.); “Kediri, Java, 1 g (K. U. Kirkaldy Coll.);.“Buitenzorg, Java, 1919 W. C. V. Heurn” ial (Calif. Acad: Sci.) The following were loaned to us for study by Dr. Wegner, Di- rector of the Zoological Museum in Bogor: “W. Java 8-900 m. Tjikadjang, Bandjarwangi 7-10, IV, 1939. M. A. Lieftinck”; 2g ¢, 39 9; C.N. Java 800 Moeria Mts. 1939 Tjolo 20-24 X, M. A. Lieftinck” 2¢ ¢,59 9. “W. Java 500m. G. Sanggaboeana 22 XII 1935 M. A. Lieftinck” 79 9; “W. Java 700m. Bandoeng 19 VII 1940 J. Olthof” 1 g ; “W. Java 700 m. Bandoeng 16 and 23 VI” 6g g, 69 9; “Palaboean ratoe 30 V 1932 Tijijolak, Lieftinck” 1¢; “W. Java 400-1000 m. Soekangegora II 1940” 1 9 ; “Wetar I- IV 1939 m. Lerai S. Bloembergen” 1 7. The following is from Dr. E. Sutter, Museum of Natural His- tory, Basel, Switzerland: “Ost Java, Lawang Djati Forests, Nglirip 200° M. E. Walsh” 1 3; “Ost Java, Lawang 1500’ M. E. Walsh” Genus PrmoMERA AMYOT AND SERVILLE 511 19; “West Java Wijnkoops Bay, M. E. Walsh” 1g, 19; “West Java, G. Molang 4290, M. E. Walsh” 1 g, 1 ¢ ; “Nord Java, Japora Distr. M. E. Walsh” 1 g¢ “West Java Soekaboemi M. E. Walsh” 1 ¢ ; Borneo: “W. Borneo. Telok Ayer F. Muir’ 2 ¢ ¢ 19, all apterous (Calif. Acad. Sci.). Ptilomera (Ptilomera) werneri Hungerford and Matsuda Plate XXXI 1958. Ptilomera werneri, Hungerford and Matsuda, Bull. Brooklyn Ent. Soc. 53:70-73. 2 plates. (Described from the Philippines). 1960. pulomene werneri, Matsuda, Univ. of Kansas Sci. Bull. 41(2):267, 539. g. 629. This was the first species we described in this genus and its distinctive characters were so surprising that we were encouraged to undertake this revision of the genus. Our description of the species was entirely adequate but followed a somewhat different outline from the one used in this paper. The types have been re- turned to the Chicago Natural History Museum and we have only a male paratype now available. Size: Male type (apterous): Length 10.9mm. to 12.85 mm.; width of head 1.68 mm. to 1.93 mm.; greatest width of body acros: mesoacetabula 3.15 mm. to 3.78 mm. Female type (apterous): Length 11.55mm.; width of hea‘ 1.76 mm.; greatest width of body across mesoacetabula 3.36 mm. Color: Reddish brown above, pale testaceous beneath. Head with two converging black lines between the eyes. Clypeus em- browned to black, antennae brownish black. Anterior margin of pronotum between the eyes black; front femur with two longitudi- nal blackish bands. Mesopleura with two black bands separated by a silvery pile. Middle and hind femora reddish brown, covered with short black setae; middle and hind tibiae darker. Structural characteristics: Relative length of antennal segments. Male holotype: Ist:2nd:3rd:4th: :137:35:40: 26. Female allotype: Ist:2nd:3rd:4th: :104:26:32:22. Actual Lengths of Leg Segments Ist tarsal 2nd tarsal Front leg Femur Tibia segment segment Male paratype. .......... 7.04 mm. 6.42 mm. 3.31 mm. 1.30 mm. Middle leg Femur Hind leg Femur Male paratype .......... 20.5 mm. Male paratype ........ 27.1 mm. Male. Length of pregenital abdominal segments:Genital seg- ments::67:50. Middle femur heavily ciliated ventrally on distal Sl, Tue UNIVERSITY SCIENCE BULLETIN Ptilomera (Ptilomera) werneri Hungerford and Matsuda Fic. 1. Male paratype, genital segments viewed from the right side show- ing the large bifurcate paramere. Fic. 2. Male paratype, genital segments as seen from below. Note the finger-like caudal projection of the pygofer. Fic. 3. Female allotype, ventral view of caudal segments. Fic. 4. Female allotype. Note the protuberance on the seventh tergite, the short connexival spines, the lack of a ventrolateral lobe on the seventh abdominal segments and the curious shape of the median caudal lobe of the seventh ventrite which is shown in Figure 8. Fic. 5. Male paratype. Note the long median lobe of the suranal plate and the very small lateral wings. The first genital segment is much larger than the second. 513 GENuS PriILOMERA AMYOT AND SERVILLE PLATE XXXI 514 ‘THe UNIVERSITY SCIENCE BULLETIN half, bare on basal half; tibia with adpressed hairs beneath. Middle and hind femora with many short black spicules above. Genital segments: Suranal plate, with median lobe broad at base but abnormally long, its tip rounded; lateral wings abnormally short. Pygofer unusually short, its dorsolateral projections directed dorsad, short and tips rounded. Dorsolateral margins of pygofer beyond dorsolateral projections parallel for a short distance and then steeply converging to form a finger like caudal tip of the pygofer. Para- mere short, stout and bifurcate at distal end. First genital seg- ment ventrally with a median longitudinal keel and with a de- pression on either side. Female. Length of thorax: Abdomen::60:70. Seventh abdomi- nal tergite plainly longer than the sixth, and with a dorsally directed protuberance on its rear margin. Connexival spines short and pointed, arising from end of connexivum. Dorsolateral lobe of seventh abdominal segment short, broad. Without ventrolateral lobe. Shape of distal median lobe of the seventh sternite as shown in Plate XXXI figs. 3, 4. Comparative notes: This is the only known species with a bi- furcated paramere in the male and the only known species with a protuberance on the seventh abdominal tergite in the female. Types: Described from holotype, allotype and two male para- types bearing the labels “Mountains w. of Lapulapu Iwahig Penal Colony, Palawan Isl. 2000 to 3000 ft. March 1-2, 1947,” and C. N. H. M. Philippines Zool. Expeditions (1946-1947) F. G. Werner leg.” The types are in the Chicago Natural History Museum except for the one male paratype in the Francis Huntington Snow Museum, KU. Genus PrILOMERA AMYOT AND SERVILLE 515 INDEX * — synonym PAGE GellonacllouBreddins +4 (RUlOmena) teen ee ae ee ee 458 aéllo cheesmanae Hungerford and Matsuda. (Ptilomera) ............ 459 GonodesuSchmncdts: A ChLOMena)) eri re yee ehy a oe ar ae eee 431 *argus Breddin (Ptilomera) (syn. of P. dromas Breddin) ............. 503 assamensis Hungerford and Matsuda (Ptilomera) .................... 421 *asbolus Breddin (Ptilomera) (syn. of P. dromas Breddin) ............ 503 breddini Hungerford and Matsuda. (Ptilomera) ..... — ............ 451 *canace Schmidt (Ptilomera) (syn. of P. cingalensis Stal) ............ 427 ceramensis Hungerford and Matsuda. n. subsp. of Ptilomera harpyia .... 470 cheesmanae Hungerford and Matsuda n. subsp. of Ptilomera aéllo Breddin, 459 chinai Hungerford and Matsuda (Ptilomera) ........................ 438 CINAAIENSis Stale CBULOTETO) Wie ee see ea arly cle Wi ON a see ei ae 427 dorcetisMBreddin « (RULOMETG) meee ee eee ah aie yeep aie npen el 501 GrOMAS ABieG Gitta CHELOTMETE)) aries tentey ote eee ee eer a 503 gressitti Hungerford and Matsuda. (Ptilomera) ..................... 446 RannaloseSchimmicts ya CUiLOMena)) ieee eee oa hereon ae See ey ic trae 493 hanpyiaa Schmidt CRUulOMmends) ew eo he ee etek an cee ee 466 harpyia ceramensis Hungerford and Matsuda. n. subsp. (Ptilomera)..... 470 himalayensis Hungerford and Matsuda (Proptilomera) ................ 414 ilactone Bred dims. © (URUOMElG) te eae See eee cee et 448 kirkaldyi Hungerford and Matsuda (Ptilomera) ...................... 444 *lachne Schmidt. (Ptilomera) (syn of P. agriodes Schmidt.) .......... 431 laclans = Breddines (RHIOMONa is sey eae eee ee ype cea ae oe 488 laticaudata Hardwicke (Evilomera)) i a hee 416 lundbladi Hungerford and Matsuda (Ptilomera) ..................... 488 maai Hungerford and Matsuda (Ptilomera) ...................5..... 476 nunikanensis Hungerford and Matsuda (Ptilomera) .................. 441 Onvasus Breedini CEHIOMETA): e 8 eh eee 497 pamphacis Breddin= (RUlLOMenG)) 9 ese ee ee ee 478 papuensis Hungerford and Matsuda (Ptilomera) .................... 454 PROPTILOMERA Hungerford and Matsuda. n. subgenus. ............ 411 TZ ET EY OWN Gl By) RIN areas pei es eae carrie EA pulVen 8) Ba arta Gan eh a BE TRtCg eon nes CeO 397 sarawakensis Hungerford and Matsuda (Ptilomera) .................. 424 Shinatisisaki a(nomensmudum) ee ee a) ere i en ye ee 403 sumbaensis Hungerford and Matsuda (Ptilomera) ................... 474 sumatranus Hungerford and Matsuda (Ptilomera) ................... 485 SUITES ONE ES SA Kliemetr ta oe ct noc earn eB OAS oc AN ea airs ce epg eee a 481 tictingsWinlers (Rulomera ule se Sas i ee ee ee 462 timorensis Hungerford and Matsuda (Ptilomera) .. ................. 436 werneri Hungerford and Matsuda. (Ptilomera) ........:°..:......... 511 ‘ oe eabeteaess hi ea THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vout. XLV] JuNE 7, 1965 [No. 6 A Morphological Study of the Autonomic Nervous System of the Opossum, Didelphis Marsupialis BY Lity S. FEenc Department of Anatomy, University of Kansas, Lawrence, Kansas TABLE OF CONTENTS PAGE Henin UC El Oiar ee teeerieeren ie tte se ate ne peers Mint NE ER orth Gaia My te oeeer uci Ae 517 Naterialsrand GNlethOaSemcee pee nantes Sn eel et Cee nate ae en chr er ee 518 ine sehoracolumbar: Division aie hoe ee ee ee eee 519 Eve. GraniosacralaDivasiom eis iis ee ca ae eer ee as 529 DISCUSSION SR ee ee nie ani nes aia wos oil Mieze ok eI bai aera eh mina el Gy Re Aa La a 534 (SHUN VE EN 0 tS corer te a SUA G a gn ier nee aN EE RN WAU ca 541 Neikeratunes@itedesc pe ee os ee ie er amen Nea nune Nay mC eRL EL Re aetna 542 Key rtorAlbbreviatliOnsi ssa ares te eee eee ae apes ea nce een a te 544 AgsstractT: The general morphology of the autonomic nervous system of the opossum, as it is revealed by gross silver staining and dissection, is described in some detail. The thoracolumbar division arises from the last cervical verte- bral level of the spinal cord cranially to the third or fourth lumbar vertebral level caudally, and is fundamentally similar in its morphology to that of other mammals. The cervical portions of the sympathetic trunks do not communicate with the vagi and communicate with the thoracic portions primarily by way of the ansa subclavia. The prevertebral ganglia of the abdominal region are poorly differentiated, one from another, and no definitive ganglion impar is recognizable. The four cranial autonomic ganglia are similar to those in other mammals, except that the submaxillary ganglion is small and diffuse’and is located im- mediately anterior to the common duct formed by the union of the sub- maxillary duct with the duct of the sublingual gland. INTRODUCTION In the past fifty or sixty years extensive morphological and phys- iological studies of the autonomic nervous system in certain mam- mals have been made by numerous investigators. A considerable (517) 21—1367 518 THe UNIVERSITY SCIENCE BULLETIN amount of the information derived from these studies concerns the human being (Kuntz, 1953, 1958); literature concerning this system in other mammals such as the rhesus monkey (Kuntz, 1933), dog (Mizeres, 1955; Kuntz, 1958) and cat (Kuntz, 1953, 1958; Reighard and Jennings, 1961) is scattered and somewhat incomplete. In- formation on the autonomic nervous system in lower mammals is fragmentary, and the only study dealing with the system in the opossum appears to be that of Broek (1908), which deals primarily with the more obvious features of the sympathetic division as it is represented in the thoracic and lumbar regions. In the absence of adequate study, it is rather commonly assumed that the autonomic nervous system is of similar morphological pat- tern in all mammals. Although the fundamental conservatism of the autonomic nervous system does provide a basis for such assump- tion, it would seem that differences of some interest might be found if certain of the relatively “primitive” mammalian forms were sub- jected to careful study. This study was initiated with the hope of supplementing the information on the autonomic nervous system in the opossum, and of comparing the system in that form with that of other better known forms. I am deeply grateful to Dr. James Hall, formerly of the Depart- ment of Anatomy, University of Kansas, for initially suggesting and guiding the bulk of this study. I should also like to express my thanks to Dr. Irwin L. Baird for his assistance in the preparation of the manuscript, to Dr. E. Raymond Hall, Director of the Museum of Natural History at the University of Kansas, who has provided material for supplementary study, and to Miss Margaret Schumann for her help in preparation of the sections used for histological study. MATERIALS AND METHODS Ten specimens of Didelphis marsupialis virginiana were used in this investigation; four specimens were collected in the vicinity of Lawrence, Kansas, and the remainder were supplied by Quivira Specialties Company, Topeka, Kansas. All were adults; six were females, and four were males. Prior to sacrifice, all specimens were kept on a diet of glucose and water for from thirty to thirty-seven days to mobilize peritoneal fat and thus facilitate dissection of nerve fibers (Richter, 1950). Following sacrifice by overdoses of pentobarbital sodium, specimens were perfused and fixed with ten percent formalin solution for at NERvous SYSTEM OF THE OPOSSUM 519 least twenty-four hours. The abdominal and thoracic cavities were then opened and the specimens were treated with 0.75% silver nitrate solution for from eighteen to twenty-four hours at room tem- perature. In some cases, nerve fibers embedded in dense connec- tive tissues were inadequately stained in this period and required local restaining to make fine nerve fibers visible. Dissections were, in all specimens, carried out with magnifications of from ten to thirty diameters provided by a stereoscopic dissecting microscope. Measurements were made directly with a scale in the case of larger structures, and with a calibrated ocular grid in the case of minute structures. Short segments cut from selected levels of the spinal cord were fixed in Bouin’s solution, embedded in paraffin blocks, sectioned at ten micra, mounted on slides and stained with Harris’ alum hema- toxylin. These served for analysis of the extent of the intermedio- lateral cell column. THE THORACOLUMBAR DIVISION As in most vertebrate animals, the thoracolumbar (sympathetic ) division of the autonomic nervous system arises segmentally from thoracic and lumbar portions of the spinal cord by way of myeli- nated fibers called white or communicating rami. These fibers originate from small multipolar neurons of the intermediolateral cell column of the central gray matter of the spinal cord, and exit from the latter in the ventral roots of the spinal nerves. Although fibers were not actually traced, microscopic observation of sections of cervical, thoracic, lumbar and sacral portions of the spinal cord show that the intermediolateral cell column in the opossum extends from the seventh cervical vertebral level to the third or fourth lumbar level. It seems reasonable to assume, therefore, that sympathetic fibers are carried in roots of all segmental nerves from the last cervical level cephalically to the third or fourth lumbar level cau- dally. It is probable, however, that the contributions from the cervical and lowermost lumbar portions of the spinal cord are ex- tremely limited, since the number of cells comprising the inter- mediolateral cell column is small in these regions. The fibers from the intermediolateral cell column accompany the spinal nerve through the intervertebral foramen, then leave the nerve and join one of the two longitudinal paravertebral sympa- thetic trunks. The latter extend along the ventrolateral aspects of 520 THe UNIVERSITY SCIENCE BULLETIN the vertebral column from the base of the skull to the sacral region. Each trunk is composed of a series of interconnected, irregular, angular or fusiform ganglia composed of aggregates of multipolar neurons with variable numbers of dendritic processes. The number or size of these ganglia is not regular or uniform in pattern, even in the two trunks of one animal; by actual comparative counting and observation in the ten animals investigated, the number of ganglia varied from six to ten in the thoracic region, four to six in the lumbar region and only one or two small swellings were visible in the sacral region. The neurons of the intermediolateral cell column are called pre- ganglionic neurons and their fibers are accordingly designated pre- ganglionic fibers. After traversing the communicating rami and joining the sympathetic trunks, preganglionic fibers may synapse with multipolar neurons in the ganglion of the same segment or continue for a distance cephalad or caudad in the sympathetic trunk before synapsing with ganglionic neurons in one of the para- vertebral or prevertebral ganglia. Axons of the ganglionic neurons extend peripherally to the structures innervated by the sympathetic nervous system and are termed postganglionic fibers. Situated behind the parietal pleura, the sympathetic trunks in the thoracic region begin at the level of the first rib and extend caudally along the ventral surfaces of the transverse processes of the thoracic vertebrae. At the level of the eleventh or twelfth thoracic vertebra, the sympathetic trunks turn slightly medially and ventrally onto the vertebral centra before leaving the thorax through the crura of the diaphragm. Each thoracic sympathetic trunk (fig. 1) includes seven to ten interconnected ganglia which, except for the first and the second ganglia, are segmentally arranged from the fourth to the tenth or eleventh thoracic vertebra. The first ganglion of the thoracic series is fusiform and is the largest of the paravertebral ganglia; it lies diagonally upon the head of the first rib in most cases, but some- times extends caudally to the level of the second rib. It appears to consist of a fusion of the first thoracic and the caudalmost cervical sympathetic ganglion and thus may be properly termed a stellate ganglion as defined by Kuntz (1953). The second ganglion, con- siderably smaller than the stellate, is usually found at the level of the third or fourth rib. The remainder of the thoracic ganglia are segmentally arranged and are located ventral to the necks of the NeERvous SYSTEM OF THE OPOSSUM 52) ribs. Each ganglion of the thoracic sympathetic trunk is connected with its corresponding spinal nerve by one or occasionally two com- municating rami. From the medial aspect of the sympathetic ganglia and intergan- glionic portions of the thoracic sympathetic trunks, numbers of slender preganglionic and postganglionic fibers arise to form the thoracic cardiac and splanchnic nerves, or pass ventrally into one of the subordinate thoracic sympathetic plexuses. Some contribute to the formation of the pulmonary plexuses and the plexus on the descending aorta. Cephalically, the thoracic sympathetic trunk is continuous with the cervical sympathetic trunk at the medial aspect of the cranial pole of the stellate ganglion (figs. 2, 3, 4). From this point a relatively small bundle of fibers courses cephalically, passes dorsal to the subclavian artery and extends into the neck along a course similar to that normally followed by the main cranial extension of the sympathetic trunk in man. Just medial to the point where this bundle springs from the ganglion, a significantly larger bundle arise, passes ventral to the subclavian artery and dorsal to the in- nominate vein and joins the dorsal branch just cranial to these vessels. The loop formed by this large ventral limb is recognized as the ansa subclavia. From it fine filaments arise and follow the subclavian artery as the subclavian plexus. From the left ansa subclavia, ventral to the subclavian artery, one to two cervical cardiac nerves arise. These soon divide into numbers of delicate filaments which travel from the base of the subclavian artery to the ventral surface of the aortic arch where they enter into the formation of a network, the superfician or ventral cardiac plexus (figs. 2, 3). Another cardiac nerve descends from the medial aspect of the caudal pole of the left stellate ganglion to the ventral surface of the aortic arch where it too breaks into fila- ments and joins the cardiac plexus. On the right side (fig. 4) a cardiac nerve derived from the caudal cervical ganglion (see be- low), or from the trunk below it, descends along the right common carotid artery to the base of the neck, where it bifurcates into two branches. The uppermost of these subdivides into many fine fibers, most of which join the ventral cardiac plexus; a few pass to the dorsal surface of the aortic arch. The lower right cervical cardiac nerve breaks down into a network which courses along the dorsal part of the superior vena cava to the dorsal surface of the aortic 522 THE UNIVERSITY SCIENCE BULLETIN arch and the ventricular wall where it joins the branches of the upper right cervical cardiac smypathetic nerve in the formation of the deepor dorsal cardiac plexus. No cardiac fibers arise from either thoracic sympathetic trunk below the level of the stellate ganglion. The ventral and dorsal cardiac plexuses communicate freely and are actually a continuous network; they are arbitrarily recognized on the basis of their positions relative to the arch of the aorta. The cardiac plexuses are the main source of the left and right pulmonary plexuses and the two coronary plexuses (figs. 3, 4). The left plumonary plexus is a small network formed largely by extension of the ventral cardiac plexus onto the left pulmonary artery. To this, a few filaments derived from the caudal end of the stellate ganglion, from the left thoracic sympathetic trunk of the second to fourth vertebral levels and from the dorsal cardiac plexus are added. The right pulmonary plexus is chiefly an extension of the dorsal cardiac plexus; there appears to be no contribution of fibers from the right thoracic sympathetic trunk below the stellate ganglion. In a similar fashion, the left coronary plexus, which follows the left coronary artery, is formed primarily by filaments from the ventral cardiac plexus with some contributions from the dorsal cardiac plexus. The right coronary plexus is formed by filaments from the dorsal cardiac plexus and a few fibers from the ventral cardiac plexus; it courses along the right pulmonary artery. Below the stellate ganglion, several delicate filaments arise from each thoracic sympathetic trunk and follow the intercostal arteries to the dorsal aorta. Here, with filaments from the opposite side, they form the aortic plexus (fig. 1), a sparse network which extends caudally along the aorta. Some of these filaments especially those in relation to the lower half of the esophagus, extend or send branches to the esophagus where they form the inconspicuous esophageal sympathetic plexus. This plexus also receives one or two fine filaments from the cardiac sympathetic nerves which arise from the ansa subclavia. From the level of the tenth to the thirteenth thoracic vertebra, each thoracic sympathetic trunk gives rise to three branches con- siderably larger than the thoracic cardiac nerves; these are the thoracic splanchnic nerves (fig. 1). The most cephalic one, the greater splanchnic nerve, is the largest and usually arises by a single root from the ganglion or interganglionic portion of the trunk at the tenth or eleventh (in one case, ninth) vertebral level. The second branch, the lesser splanchnic nerve, is slightly smaller than the NeERvous SYSTEM OF THE OPOSSUM 523 greater splanchnic nerve and is derived by a single root from the twelfth or thirteenth vertebral level of the thoracic sympathetic trunk. The third or least splanchnic nerve arises a few millimeters below the origin of the lesser splanchnic nerve. More variable than the other two, it originates at either the thirteenth thoracic or first lumbar vertebral level, or from the trunk between them. In some cases, the nerve has but a single root; in others, it arises from one main root which is supplemented by smaller rootlets from adjacent portions of the trunk. Along each splanchnic nerve, especially the greater and lesser, ganglion-like swellings are often observed. Similar structures are described in man and cat (Kuntz, 1956) and are designated as ganglia included in the splanchnic nerves. The splanchnic nerves are, in most cases, unilaterally interconnected by branches and they frequently divide and/or combine in (apparently) a completely in- consistent manner before leaving the thorax by piercing the crus of the diaphragm. From its origin anterior to the subclavian artery at the junction of the ansa subclavia and the more dorsal sympathetic trunk, the cervical sympathetic trunk extends cephalically, in the groove be- tween the trachea and the esophagus, along the ventral surface of the transverse processes of the seventh to the second cervical verte- bra and is separated from the latter by the longus capitis and the Jongus collimuscles. At its origin, the cervical sympathetic trunk lies lateral and dorsal to the carotid artery and the vagus nerve. Almost immediately after its origin it turns medially and dorsally and, throughout the remainder of its course, each cervical sym- pathetic trunk maintains a position immediately dorsal to the carotid artery and medial to the vagus nerve and jugular vein. Each trunk is enclosed in a heavy sheath of connective tissue which is tightly bound to the connective tissue enclosing the carotid artery, jugular vein and vagus nerve. Unlike the arrangement seen in the thoracic region, no communi- cating rami were found to connect the cervical sympathetic trunk with the cervical spinal nerves. Communicating rami to the last five cervical spinal nerves of each side arise from the cranial pole of the corresponding stellate ganglion. That to the last cervical spinal nerve arises individually from the lateral side of the cranial end of the stellate ganglion; communicating rami to the fourth, fifth, sixth and seventh cervical spinal nerves arise as branches from a 594 THe UNIVERSITY SCIENCE BULLETIN large common trunk from the cranial pole of the stellate ganglion (figs. 2, 3, 4). The rami to the posterior cervical nerves are con- siderably larger than those to the more anterior nerves; the ramus to the last cervical spinal nerve is reasonably large and well-defined but, anterior to this, each communicating ramus is found to be progressively smaller until, at the fourth cervical spinal nerve, only a delicate filament connects with the sympathetic nervous system. No communicating rami to the first, second and third cervical spinal nerves were found in any of the specimens dissected. The three ganglia present on each cervical sympathetic trunk (figs. 2, 3, 4) are not segmentally arranged as are those in the thoracic region. The most cephalic of these, the cranial cervical sympathetic ganglion, is fusiform, is five or six millimeters in length and is situated dorsomedial to the carotid artery just caudal to the bifurcation of that vessel; the vagus nerve is contiguous with the lateral margin of the cranial cervical sympathetic ganglion. From the cranial end of this ganglion, four or five large branches arise, pass to the internal carotid artery and follow it into the skull. These branches constitute the origin of the carotid sympathetic plexus. This plexus extends as a heavy network along the carotid arteries and is the only apparent source of sympathetic innervation in the head region. A few millimeters caudal to the foramen lacerum a small bundle of nerve fibers leaves the carotid plexus as the (great) deep petrosal nerve and immediately joins the great superficial petrosal nerve (which originates at the geniculate ganglion of the facial nerve and is embedded in a groove on the dorsal surface of the sphenoid bone) to form the nerve of pterygoid canal (Vidian nerve). The latter proceeds forward through the pterygoid canal, traverses the inferior orbital fissure and enters the sphenopalatine fossa to terminate at the sphenopalatine ganglion (see below). The remainder of the carotid plexus continues forward into the skull with the internal carotid artery. Fine networks of filaments follow each branch of the carotid artery, but could not be traced for sig- nificant distances along these branches. Although an intensive study of the sympathetic innervation of the head was not conducted, the above observations on cranial sym- pathetic distribution in the opossum disclose no major differences from conditions described in the cat (Reighard and Jennings, 1961 ) and man (Kuntz, 1953). It seems reasonable to assume that the whole pattern of sympathetic distribution in the head region of the opossum may be much like that in cat and in man. Nervous SYSTEM OF THE OPpossuM 595 The second ganglion, the middle cervical sympathetic ganglion (figs. 2, 3, 4) is a slender spindle-shaped body approximately twice as thick as the interganglionic trunk and is situated midway between the stellate and cranial cervical sympathetic ganglion. The third ganglion is merely a slight swelling of the cervical sympathetic trunk immediately anterior to its origin. Since it corresponds in position to the caudal cranial sympathetic ganglion of the dog (Mizeres, 1955), it seems logical that this ganglion be recognized by that name. In some specimens, this ganglion appeared to be absent on one side, but in none of the specimens was it missing bilaterally. The sympathetic trunks in the lumbar region (fig. 5) are direct continuations of those in the thoracic region. Caudal to their entries into the abdominal cavity through the crura of the diaphragm dorsal to the thoracic splanchnic nerves, the trunks incline medially and course caudally, adjacent to the dorsal aorta and partially under the cover of the psoas muscles. As they approach the sacral region, the lumbar sympathetic trunks pass dorsal to the aorta, approach the midline and diminish in size; here they become interconnected by delicate filaments. Each trunk exhibits four or five ganglia which are relatively flatter than those of the thoracic region and are ir- regular in size and in shape. As in the thoracic region, each lumbar sympathetic trunk is connected with the lumbar spinal nerves by communicating rami. Medially, each ganglion of the lumbar sympathetic trunk gives rise to several fine branches, lumbar splanchnic nerves (fig. 5), which contribute to either prevertebral ganglia or plexuses in the abdominal cavity. The interganglionic portions of the trunk also give off a few fine filaments which tend to follow the same distribu- tion as the splanchnic nerves from the ganglia or combine with them a short distance from the trunk. The splanchnic nerves from the first lumbar sympathetic ganglion contribute chiefly to the celiac ganglionic complex (see below), but some filaments are directed to the cranial mesenteric ganglion (see below). Frequently, the splanchnic nerves from the first lumbar sympathetic ganglion unite with the descending thoracic splanchnic nerves prior to their termi- nation in the celiac ganglionic complex. Lumbar splanchnic nerves from the second ganglion end partly in the cranial mesenteric gan- glion, but most of them contribute to the intermediate mesenteric or the aortic plexus (see below) on the ventral surface of the ab- dominal aorta; the latter filaments are supplemented by splanchnic 526 THe UNIVERSITY SCIENCE BULLETIN nerves from the third ganglion and some from the fourth ganglion. The majority of the splanchnic nerves from the fourth ganglion terminate chiefly in the caudal mesenteric ganglion (see below), which is located in the mesentery anterior to the bifurcation of the aorta. Splanchnic filaments from the fifth lumbar ganglion follow the hypogastric artery to enter into the formation of a plexus of the same name (see below). The major prevertebral ganglia of the abdominal region (fig. 5) are the celiac, cranial mesenteric, intermediate mesenteric and the caudal mesenteric ganglia. From these ganglia, numerous fine fila- ments arise and form extensive prevertebral plexuses. The bilateral celiac ganglia are large, complex, flattened structures of irregular shape and are located in an area bounded by the common trunk of the celiac artery, the cranial mesenteric artery, the renal arteries, and the aorta. Each celiac ganglion is arbitrarily separable into a number of smaller ganglia. At the origin of the renal artery, a small star-shaped gangliform mass is loosely attached to the anterolateral side of the celiac ganglion and may be designated as the aorticorenal ganglion; from this, filaments arise and constitute the main source of the renal plexuses along the renal arteries. Caudally, each celiac ganglion is extended and, on the ventral surface of the aorta pos- terior to the common trunk of the celiac and cranial mesenteric arteries, this extension fuses with that of the opposite side to form the irregular cranial mesenteric ganglion. It must be emphasized that recognition of these three ganglia in the opossum is rather arbitrary since they appear as one continuous structure which sur- rounds the common trunk of the celiac and cranial mesenteric arteries dorsally and posterolaterally as a semicircular band. From the cranial end of each celiac ganglion, a few filaments arise and, as the delicate phrenic plexus, follow the phrenic artery to the diaphragm. Several fibers arise from the lateral side of each celiac ganglion adjacent to the renal artery and join fibers from the aor- ticorenal ganglion to form the renal plexus. This plexus courses along the renal artery as a delicate network and enters the kidney. At the hilus of that organ, a few fine filaments leave the renal plexus and follow the ureteric artery and ureter to the dorsal surface of the bladder. From the ventral surface and anterolateral margin of each celiac ganglion, numerous fibers pass directly to the adrenal gland as the adrenal plexus; these accompany the artery to the ad- renal gland, which is located medial to the cephalic end of the kidney. On the right side, the adrenal gland is often attached to the Nervous SYSTEM OF THE OPOSSUM 527 tip of the right lobe of the liver and to the portal vein by dense connective tissue; through this some filaments of the adrenal plexus pass to the portal vein and follow the latter to the liver. The celiac and cranial mesenteric arteries in the opossum orig- inate from a common trunk which divides into two branches (fig. 5). One branch turns to the left as the splenic artery, which gives rise to a common stem for the left gastric and the left phrenic arteries a short distance from its origin, then courses through the body and tail of the pancreas to the spleen and greater curvature of the stomach. The second branch of the original trunk gives rise to the hepatic artery and its branches (gastroduodenal and pancreatic arteries), to the crainial mesenteric artery and to a descending caudal mesenteric artery. The right celiac ganglion contributes fibers primarily to the formation of the hepatic plexus and a few fibers to the cranial mesenteric plexus. The hepatic plexus is a moderately dense network which courses along the artery and its branches to reach the liver, gall bladder, head of the pancreas, lesser curvature of the stomach and duodenum. Additionally, the right celiac ganglion contributes a few fibers to the cranial mesenteric plexus (see below). The left celiac ganglion contributes heavily to the formation of the splenic plexus, which follows the splenic artery and its branches to the spleen, to the body and the tail of the pancreas and to the greater and lesser curvatures of the stomach. Several fibers leave the caudal margin of the left celiac ganglion and contribute to the cranial mesenteric plexus. Although direct connections between the two celiac ganglia are not obvious, fine fibers from both sides do mingle in a network at the bases of the cranial mesenteric and hepatic arteries. The cranial mesenteric ganglion (fig. 5) is a semilunar band which connects the slender posterior ends of the celiac ganglia and is situated on the ventral surface of the aorta immediately posterior to the common trunk of the celiac and cranial mesenteric arteries. This ganglion is the main source of the cranial mesenteric plexus, a dense network of relatively course fibers associated with the cranial mesenteric artery. As noted above, a few fibers from each celiac ganglion descend and contribute to the plexus near its origin. A large portion of the cranial mesenteric plexus follows the main artery and its branches in their distribution to the jejunum, ileum, and cecal region of the colon. The remainder of the plexus follows the caudal mesenteric artery to the greater part of the colon and to the 528 THe UNIVERSITY SCIENCE BULLETIN rectum. A few fibers arise from the cranial mesenteric ganglion and enter the right and left renal plexuses, and several course fibers arising from the caudal end of the ganglion contribute to the forma- tion of the intermediate mesenteric plexus. The latter is a rather dense network containing several scattered ganglia and long heavy parallel bundles of fibers which extend between the cranial and caudal mesenteric ganglia (see below). This plexus is located on the ventral surface of the aorta and extends from the caudal end of the celiac ganglionic complex to the bifurcation of the aorta. At the origins of the gonadal arteries the plexus contains the small paired gonadal ganglia. Each of these is connected anteriorly with the intermediate mesenteric plexus by a large bundle of fibers and posteriorly with the caudal mesenteric ganglion by a few delicate filaments; each also receives several fine fibers from the third lumbar splanchnic nerve. Filaments from the gonadal ganglia, sup- plemented by a few fibers directly from the intermediate mesenteric plexus, form the gonadal plexuses, which accompany the arteries to the gonads. These plexuses also receive contributions from the ureteric plexuses as they descend. At the bifurcation of the aorta, the intermediate mesenteric plexus gives rise to two sets of fibers; each of these follows a common iliac artery and its branches and, on each side, the major part of these fibers accompanies the hypogastric artery as the hypogastric plexus (fig.5). The latter passes posteriorly to the side of the rectum and region of the neck of the bladder in the male, and to the side of the rectum and vagina in the female. The hypogastric plexuses are the main sources for the formation of the pelvic plexuses, which are provided with supplementary contributions of fibers from the sacral sympathetic trunks and sacral spinal nerves at the first and second sacral vertebral levels (figs. 6, 7). The pelvic plexuses are tightly packed networks situated ventrolateral to the rectum. Each has a large irregular pelvic ganglion in its center and, in some specimens, several tiny ganglia are found scattered in the plexus near the neck of the urinary bladder. These plexuses give rise to a number of poorly-defined subordinate plexuses named according to the organs they innervate. These pass ventrally to the urinary bladder, prostate gland and urethra in the male (fig. 6), and to the urinary bladder, uterus and vagina in the female (fig. 7). Just caudal to the level of origin of the gonadal arteries, approxi- mately twenty millimeters anterior to the bifurcation of the aorta, the intermediate mesenteric plexus sends a few fibers into the root Nervous SYSTEM OF THE OPpossuM 529 of the mesentery; these pass to the caudal mesenteric ganglion (fig. 5), an irregular, flattened, bilobed structure approximately twice as large as the gonadal ganglion. It is located within the mesentery ventral to the aorta and to the right of the mid-sagittal plane. Anteriorly, this ganglion receives a few long bundles of fibers from the celiac ganglionic complex and, dorsally or laterally, it receives the splanchnic nerves of the fourth lumbar ganglion. From the caudal mesenteric ganglion, a relatively large trunk passes to the gonadal ganglion of each side. A few fibers leave the caudal end of ganglion and pass to the gonadal artery of each side to supple- ment the gonadal plexuses. In addition, the caudal mesenteric ganglion gives off fibers which course through the mesentery to the caudal mesenteric artery and, by following the latter, reach the descending colon and the rectum. THE CRANIOSACRAL DIVISION The craniosacral or parasympathetic division of the autonomic system arises from two discontinuous regions, the cranial and the sacral portions of the central nervous system. The functions of this division have been clearly shown to be antagonistic to those of the thoracolumbar division in various viscera of different animals (Bab- kin, 1946). The cranial portion is derived from certain nuclei located in the gray matter of the midbrain and brain stem and com- posed of preganglionic parasympathetic neurons. From these neu- rons, preganglionic fibers pass peripherally in the third, seventh, ninth, and tenth cranial nerves to one of four cranial parsympa- thetic ganglia (ciliary, sphenopalatine, submaxillary, or otic gan- glion) or to one of the numerous small ganglia in the walls of the structures innervated. In these ganglia, the preganglionic fibers make their synapses with ganglionic neurons and these, in turn, send postganglionic fibers to the effectors. The sacral portion arises from the medial autonomic nuclei of the gray matter of the spinal cord in the sacral and, probably, lowermost lumbar regions. As pregan- glionic fibers, axons from cells of these nuclei proceed in part as pelvic nerves to the pelvic ganglion where they synapse with neurons which send postganglionic fibers peripherally. Other preganglionic fibers pass directly to certain of the pelvic viscera and there synapse with small intramural ganglia. Preganglionic parasympathetic fibers in the cranial region are im- possible to trace, since they are incorporated into the cranial nerves. The four cranial parasympathetic ganglia are, however, identifiable 530 THe UNIVERSITY SCIENCE BULLETIN in all specimens, and sufficient of their connections are visible to permit extrapolations concerning sources and courses of parasym- pathetic fibers. Ciliary ganglion: The rostralmost cranial autonomic ganglion is the ciliary ganglion, a tiny fusiform structure approximately one millimeter in length and one and one half millimeters in width (fig. 8). It is located to the lateral side of the optic nerve, between that structure and the tendons of the inferior rectus and lateral rectus muscles, and lies about eight millimeters from the optic foramen. According to Kuntz (1953), this ganglion is ontogenetically related to the oculomotor and ophtha!mic nerves. As in the cat (Reighard and Jennings, 1961) this ganglion in the opossum is attached to the inferior division of the oculomotor nerve by the short ciliary root, a bundle of preganglionic fibers from the ipsilateral Edinger-West- phal nucleus; some of these fibers merely course through the gan- glion but most of them synapse in the ganglion. On its upper border, the ciliary ganglion receives, as the long ciliary root, a few fine filaments from the ophthalmic branch of the trigeminal nerve; these fibers are probably sensory and originate from neurons of the semilunar ganglion. This differs greatly from that in the cat (Christensen, 1936). In company with the long and short roots, some postganglionic sympathetic fibers from the internal carotid and cavernous plexuses probably reach and traverse the ciliary gan- glion. Two to five fine filaments from the cranial end of the ciliary ganglion extend along the optic nerve as the short ciliary nerves and enter the posterior surface of the eyeball. Sphenopalatine ganglion: The sphenopalatine ganglion (fig. 9) is the largest cranial parasympathetic ganglion. It is a flattened fusiform structure two millimeters in width and six millimeters in length, and is situated in the external pterygoid fossa approximately six millimeters caudal to the sphenopalatine foramen. Medially, this ganglion is in intimate contact with the pterygoid bone; laterally it is covered by the inferior rectus muscle and is adjacent to the infraorbital branch of the maxillary nerve. Three or four short (sensory ) fibers arise from the latter and enter the lower margin of the ganglion. Posteriorly, the sphenopalatine ganglion receives the nerve of the pterygoid canal. The latter is formed by the great superficial petrosal nerve (from the facial nerve), which proceeds anteriorly in a groove on the dorsal surface of the sphenoid bone and, a few millimeters posterior to the foramen lacerum, unites with NeRvous SYSTEM OF THE OPossuM 581 the deep petrosal nerve (see above). From this union, the nerve traverses the pterygoid canal and enters the caudal margin of the ganglion as a large trunk. From the anterior tip of the ganglion, a large trunk springs out and splits immediately into two branches, the posterior nasal nerves, which course through the sphenopalatine foramen to the nasal cavity, to the sinuses, and to the hard palate. Just caudal to the origin of the above nerves, a smaller branch arises from the upper margin of the ganglion as the major or great palatine nerve to the hard palate. From the medial surface and the inferolateral margin of the ganglion, two or three fine filaments course to the soft palate as the minor or small palatine nerves by traversing the lesser or accessory palatine foramina. From the upper border of the spheno- palatine ganglion, a small bundle of fibers arises near the anterior limit and travels upward to the lacrimal gland. A few filaments from the posterior end of the ganglion follow the ethmoidal branch of the ophthalmic artery to the ethmoidal air cells. Submaxillary ganglion: In the opossum, the sublingual and sub- maxillary glands are located together at the posterior border of the masseter muscle below the parotid gland and lie at the junction of the lower jaw and anterior neck. The duct of each gland extends forward a few millimeters from the anterior border of the gland; the two then join to form a common duct which pierces the digastric and mylohyoid muscles to reach the mouth. This duct courses a few millimeters below the lingual branch of the mandibular nerve. The submaxillary ganglion (fig. 10), an irregular structure less than one millimeter in diameter, is situated ventral to the sub- maxillary duct immediately posterior to its union with the duct of the sublingual gland; in some cases, the ganglion is related to the union itself or to the posterior part of the common duct. Anteriorly, this ganglion receives fine filaments which descend to the common duct from the lingual nerve directly above, then travel posteriorly for some distance along the duct to reach the submaxillary ganglion. Filaments leave the ganglion posteriorly and medially to enter the adjacent salivary glands. Otic ganglion: The otic ganglion (fig. 11) is a flattened, minute structure (in some cases, it is bilobed) approximately one and one- half millimeters in length; it is located medial to the anterior branch of the mandibular nerve immediately below the foramen ovale. The ganglion is in contact anteriorly with the posterior margin of the 5382 THe UNIVERSITY SCIENCE BULLETIN internal pterygoid muscle and posteriorly with the tensor veli pala- tini muscle. Caudally, the otic ganglion receives a small bundle of fibers from the glossopharyngeal nerve; anteriorly it is connected with a branch from the anterior division of the mandibular nerve, which innervates the internal ptrygoid muscle and the nerve of the pterygoid canal by two fine filaments. From the ventral border of the ganglion, three or four delicate filaments are sent to the tensor veli palatini muscle, which separates the ganglion from the auditory tube. Laterally, the otic ganglion gives off two fine branches; one enters the mandibular nerve next to the ganglion the other follows an artery to the parotid gland. No contribution to the auriculo- temporal nerve can be found. Vagus nerve: The course of the vagus nerve in the opossum (figs. 1, 3, 4) is similar to that described in the cat (Reighard & Jennings, 1961). Each vagus nerve exists from the skull through the jugular foramen, then, as a large trunk, descends lateral to the trachea to the point of bifurcation of the latter. It follows the esophagus through the diaphragm and enters into the formation of the celiac and other subordinate plexuses in the abdominal cavity. From the vagus nerve, numerous branches arise at various levels of its course and enter into the autonomic plexuses above the dia- phragm. In the cervical region, the courses of the two vagi are virtually identical. Immediately below the jugular foramen, lies the nodose ganglion, which is ovoid, approximately seven or eight millimeters in length, and situated in the carotid sheath anterolateral to the cranial sympathetic ganglion and carotid artery. From the medial margin of the caudal part of this ganglion, the cranial laryngeal nerve arises and crosses ventral to the cranial sympathetic ganglion and carotid artery in its course to the larynx. As it crosses the artery, the laryngeal nerve gives off a cardiac nerve (figs. 3, 4), which first courses posteriorly along the ventral surface of the com- mon carotid artery; approximately midway down the neck the cardiac nerve passes to the dorsal surface of the same artery and courses posteriorly to the arch of the aorta. There it breaks into numerous delicate filaments which enter the dorsal cardiac plexus together with the filaments of the same nerve of the opposite side. Beyond the origin of the cranial laryngeal nerve, the vagus nerve courses posteriorly for a short distance, then turns somewhat medially to lie laterally adjacent to the carotid artery and ventral to the cervical sympathetic trunk. It maintains these relationships NERVOUS SYSTEM OF THE OPpossuM 533 in the rest of its course in the cervical region, but becomes more superficial as it approaches the base of the neck. Beyond the origins of the common carotid arteries the courses and distribution of the two vagi differ. At the base of the neck the left vagus nerve proceeds posteriorly, between the left common carotid and left subclavian arteries, first on the ventral surface of the arch of the aorta, then courses dorsal to the left innominate vein, pulmonary vessels and the left bronchus. Beyond this level, the nerve continues posteriorly on the ventral surface of the esophagus and enters the abdominal cavity with that structure. Immediately caudal to the diaphragm, the left vagus nerve breaks down into numerous branches; most of these enter the splenic and gastric plexuses along the arteries of the same names; the rest enter the cranial end of the left celiac ganglion. Several branches arise from the thoracic portion of the left vagus nerve. Immediately below the arch of the aorta, it gives off a rel- atively large branch, the recurrent laryngeal nerve, which turns craniad, dorsal to the arch, and courses anteriorly adjacent to the trachea. In this portion of its course, the recurrent nerve sends several small branches to the dorsal surface of the trachea and to the esophagus. Approximately five millimeters caudal to the larynx, the recurrent laryngeal nerve branches into five to seven fine fila- ments which enter the larynx separately. Caudal to the origin of the recurrent laryngeal nerve, the left vagus nerve gives rise to four or five fine fibers; two or three of these enter the dorsal and ventral cardiac plexuses; the other two course to the ventral surface of the left pulmonary artery and bronchus to enter the left pulmonary plexus. Beyond the pulmonary root, the left vagus nerve also gives off several fine filaments which join the esophageal plexus. The course of the right vagus nerve through the thoracic cavity differs somewhat from that of the left vagus nerve. It passes ventral to the right subclavian artery, dorsal to the right innominate vein, right bronchus and pulmonary vessels. Beyond the root of the lung, the right vagus nerve courses posteriorly on the dorsal surface of the esophagus and follows the latter into the abdominal cavity. After entering the abdominal cavity, the right vagus nerve, like the left, branches into numerous fine fibers. The majority of these contribute to the hepatic and mesenteric plexuses related to the arteries of the same names, but a small part of the fibers enters the anterior margin of the right celiac ganglion. The right vagus nerve gives off a recurrent laryngeal nerve immediately caudal to the right 534 THe UNIVERSITY SCIENCE BULLETIN subclavian artery and passes anterolaterally adjacent to the trachea. The light recurrent laryngeal nerve, like the left, branches into several fine filaments shortly before its entrance into the larynx. Caudal to the origin of the recurrent nerve, the right vagus nerve gives off one or two fine filaments which follow the right sub- clavian artery and the right innominate vein to the arch of the aorta and enter the dorsal and ventral cardiac plexuses. Dorsal to the right pulmonary root, two fine filaments arise from the right vagus, spiral ventrally around the bronchus and pulmonary vessels and join the right pulmonary plexus. On the right dorsal surface of the esophagus, several delicate filaments arise from the right vagus nerve and contribute to the dorsal part of the esophagus plexus. The sacral portion of the craniosacral division of the autonomic nervous system is comprised solely of the fibers originating from certain of the medial parasympathetic nuclei of the spinal cord; these fibers exit from the cord in company with the first and second sacral nerves. Shortly after the sacral nerves are outside of the spinal canal, these fibers diverge as two or three small bundles, the pelvic nerves, which proceed to the pelvic ganglia on either side of the rectum. From the pelvic ganglia, numerous delicate fila- ments arise and mingle with sympathetic fibers in the pelvic plexus and its subordinate parts. DISCUSSION Although knowledge of the autonomic nervous system has been derived mainly from the study of mammals, fragmentary informa- tion concerning this system in lower chordates reveals that the autonomic nervous system is, phylogenetically, a rather primitive system. It can be traced as a diffuse enteric network, apparently independent of the central nervous system, in the regions of the heart, pharynx and gut in the balanglossids or acorn worms, chor- dates presumed to be not far removed from early chordate stem stock (Bullock, 1945; Dawydoff, 1948; cited by Nicol, 1952; Boeke, 1949). In Amphioxus visceral fibers are found in the dorsal roots of the segmental spinal nerves; in forms more advanced than this, visceral fibers occur in both dorsal and ventral roots of the spinal nerves. This tends to indicate that the visceral network is the primitive basis of the autonomic nervous system, and that its connec- tion with the central nervous system is secondary. Within the chordates, the autonomic nervous system shows addi- tional evidence of progressive elaboration of complexity in mor- Nervous SYSTEM OF THE OPpossuM 535 phology and physiology, and some indications of diversification of organization and specialization are clearly shown. In the essentially brainless Amphioxus there is no parasympathetic system; visceral fibers in the dorsal roots of the spinal nerves reach the gut, gill pouches and other visceral structures after traversing the dorsal body wall (Goodrich, 1958). In cyclostomes early representation of both sympathetic and parasympathetic innervation of the viscera is present. The cardiac plexus is composed solely of contributions from the vagus nerve, but the latter has no functional effect on cardiac activity. Scattered interconnected cells or ganglia are sparsely spread along the cardinal veins from the heart to the level of the anus (Kuntz, 1911b; Kraus, 1923; Nicol, 1955; Goodrich 1958). Sympathetic ganglia in elasmobranchs are segmentally but loosely arranged along the vertebral column, and there is no sym- pathetic encroachment in the head region. In teleosts, an organized sympathetic cord with regular ganglia interconnected by longi- tudinal strands is characteristic, and sympathetic fibers pass into the cranial region ( Young, 193la; Young, 1933a; Hirt, 1934; Daniel, 1934). In amphibians, particularly in anurans, a well-organized basic pattern of the autonomic nervous system, with clearly defined sympathetic and parasympathetic divisions, is found. Work con- cerning the autonomic nervous system in reptiles and birds (Terni, 1931; Hirt, 1935; and others) indicates that this system is rather similar to that of mammals in its general basic organization. It is distinctly divided into craniosacral and thoracolumbar divisions, and limited information indicates that its function may approach, in diversity and complexity, that found in mammals. Functionally, the sympathetic and parasympathetic systems seem to exert increasing numbers of discrete effects from elasmobranchs to mammals. In elasmobranchs the vagi have an inhibitory effect on cardiac activity, and sympathetic fibers exert a motor effect on abdominal viscera (Lutz & Wyman, 1932; Young, 1933a; Nicholls, 1934). In teleosts, the sympathetic system also exerts influence upon the eyes, chromatophores, swim-bladder gland and blood ves- sels, in addition to the functions present in elasmobranchs. The control of the sympathetic system over the heart and intestinal glands is gradually extended in amphibians (Miller & Liljestrand, 1918; Young, 1931). In mammals, the autonomic system has been considered to play important roles in maintenance and regulation not only of visceral activities, but also in reproduction, homeo- stasis, and in stress. Its widespread influence in the human being 536 Tue UNIVERSITY SCIENCE BULLETIN is clearly shown in many excellent studies, such as those of Cannon (1936), Gellhorn (1943), Kuntz (1951) and others. It should be pointed out, however, that experimental studies of the autonomic nervous system in lower mammals and other vertebrates are ex- tremely few, in comparison to the intensive work which has been done with higher mammals. It may well be that, despite its dif- fuse morphologic organization in lower forms, the autonomic nerv- ous system subserves many more discrete functions than those presently recognized. This survey of the morphology of the autonomic nervous system in the opossum indicates that its organization (fig. 12) is quite con- sistent with that which might be expected in a generalized mammal. The basic pattern is somewhat advanced over that in reptiles, and is fundamentally like that of other mammals. The sympathetic system appears as two longitudinal ganglionic chains which extend from the base of the skull to the sacral region; each chain connects with the segmental spinal nerves by communicating rami. Cranially, each trunk gives rise to fibers which follow blood vessels to supply structures in the head. Numerous filaments arise from the trunks to innervate the viscera directly or to enter prevertebral ganglia. The parasympathetic division is represented by fibers of the third, ~ seventh, ninth and tenth cranial nerves and their related ganglia, and by the pelvic nerves derived from the first and second sacral spinal nerves. The similarity of this system in the opossum to that in both lower and higher vertebrates is, perhaps, an indication that this is a primitive and conservative system well adapted to make necessary internal adjustments to changes in both internal and ex- ternal environment. There are, in the opossum, numerous slight variations of the autonomic nervous system from conditions described in other mam- mals. Although the functional significance of these variations can- not be evaluated at this time, it is worthwhile to consider them in some detail. Unlike the cat, rhesus monkey and human being in which the cervical spinal nerves are connected with the sympathetic trunks by communicating rami derived from the cranial cervical, middle cervical and lower cervical or stellate ganglia, the lower five cervical spinal nerves (fourth, fifth, sixth, seventh and eighth) in the opos- sum receive communicating rami that originate from the cranial pole of the stellate ganglion. These communicating rami arise from a common root which travels with the vertebral artery. Communi- Nervous SYSTEM OF THE OPpossuM om cating rami of similar origin and pattern are described in some mammals by Broek (1908) and Hirt (1935); these workers designate the common root of these rami the vertebral nerve, a term which seems appropriate to apply to the structure in the opossum. Micro- scopic study indicates that no intermediolateral cell column is pres- ent in the cervical region of the spinal cord, so it would seem that the communicating rami arising from the vertebral nerve are com- posed of postganglionic fibers arising from the upper thoracic levels of the cord. Although no communicating rami were found to join the upper three cervical spinal nerves, it is probable that they are present and that they arise from the cranial cervical sympathetic ganglion as in certain other mammals; Broek (1908) figures and describes this pattern in several forms, and communicating rami to all spinal nerves of the cervical region have been found to be present in representatives of all vertebrate groups from teleosts to mammals. Work by Terni (1931), Hirt (1935) and others indi- cates that, in reptiles and birds, the cervical sympathetic trunk con- sists of superficial and deep portions; in some cases in birds the former may fuse with the vagus nerve, and the latter, the vertebral nerve, always travels with the vertebral artery and is connected with the cervical spinal nerves by communicating rami. The origin and pattern of the vertebral nerve in the opossum and, according to Broek, in some other mammals (unfortunately unnamed), appears to be more comparable to that in reptiles and birds than to the single cervical sympathetic trunk in the cat, rhesus monkey and human begins. It would be attractive to interpret this as an indica- tion of a relatively primitive condition of the autonomic nervous system in the cervical region, but data available are inadequate to warrant such a conclusion. The sympathetic trunk and the vagus nerve are separate in the opossum, as they are in the rhesus monkey (Kuntz, 1933) and human being (Kuntz, 1953). Broek (1908) classifies mammals in which the sympathetic trunk is separate from the vagus nerve in the cervical region as members of one of two major groups; the second group includes those mammals in which the cervical sym- pathetic trunk is fused intimately with the vagus nerve, as is the case in the dog and cat. This classification seems to be question- able, since many variations of relationship exist between the ex- tremes of total fusion and complete separation. For example, con- nections by fine branches between the cervical sympathetic trunk and the vagus nerve have been found at various levels in the rhesus 538 THe UNIVERSITY SCIENCE BULLETIN monkey and human beings; in some cases, partial fusion of the two trunks may be present in the lower part of the cervical region in the cat. In the opossum, however, there is no connection between the sympathetic trunk and the vagus nerve. This complete separa- tion of the sympathetic trunk from the vagus nerve in the cervical region of the opossum may be another indication of less complexity or of a more primitive pattern of the autonomic nervous system than is found in other mammals. In comparing the cardiac plexus of the opossum to that in the cat, dog, rhesus monkey and human being, it is notable that the cardiac nerves derived from the sympathetic trunk and vagus nerve are fewer in number in the opossum, and that the sympathetic cardiac nerves are derived chiefly from the stellate ganglion and the trunk immediately above the stellate ganglion. There are no sympathetic cardiac nerves derived from the cranial or middle cervical ganglia or from the cephalic portion of the cervical sym- pathetic trunk, as is the case in the cat, rhesus monkey and human being. In the latter forms mentioned, the cardiac plexus is formed by two or three sympathetic cardiac nerves derived from each thoracic sympathetic trunk, plus three or more cardiac fibers con- tributed by each cervical sympathetic trunk. In the opossum only one cardiac fiber is contributed by each of the cervical sympathetic trunks, and three sympathetic cardiac nerves are contributed by the stellate ganglia, one from the right and two from the left; no cardiac nerves arise from the sympathetic trunk below the level of the stellate ganglion. Unlike the condition in the human being, in which each vagus nerve gives off two cardiac nerves in both cervical and thoracic regions, each vagus nerve in the dog, rhesus monkey and opossum gives off only one cardiac nerve in the cervical region and, in the case of the cat, it gives off none; in the thoracic region, the vagus nerve usually gives rise to two, three, or sometimes more cardiac branches in all of these forms. The relative sparseness of the cardiac plexus and the limited number of small cardiac nerves contributing to it lead one to speculate concerning the amount of control the autonomic nervous system exerts over the heart in the opossum. It would be interesting to determine whether or not such control is comparable to that found in more active and vigorous forms. The source of the esophageal plexus in the opossum is similar to that in the cat, dog, and human being in that it is derived from the sympathetic trunks and vagi of the cervical and thoracic regions, Nervous SYSTEM OF THE OPossuUM 539 but the number of fibers contributed to the plexus is smaller and the network is inconspicuous in the opossum in comparison to that in the other forms mentioned. In the region of the subclavian artery in the cat and human being the thoracic sympathetic trunk continues into the cervical region dorsal to the subclavian artery, and slender loop, the ansa subclavia, passes ventral to the artery (Kuntz, 1953; Reighard and Jennings, 1961). In contrast to the condition in these mammals, the ansa subclavia in the opossum is large and appears to form the major connection between the thoracic and cervical parts of the sym- pathetic trunk. It gives rise to the sympathetic cardiac nerves and to fibers which accompany the subclavian artery of the same side. The size of the ansa subclavia may be the result of the manner or sequence in which the stellate ganglion is formed in relation to the subclavian artery, or to a more complex developmental pattern, since fibers contributed to the cardiac and esophageal plexuses arise primarily in this region. The levels at which the thoracic splanchnic nerves originate from the sympathetic trunk vary considerably in mammals. In the human being the splanchnic nerves are derived from the fifth to the twelfth thoracic vertebral level (Kuntz, 1956). In the cat, they are described as originating from the twelfth thoracic to the first lumbar vertebral level (Kuntz, 1956). The splanchnic nerves arise from the thirteenth thoracic and the first lumbar vertebral levels in the dogs and rhesus monkey (Kuntz, 1933; Mizeres, 1955); these splanchnic nerves, in most cases, unite as one single trunk to enter the abdominal region. In the opossum, the splanchnic nerves are derived from the tenth or eleventh thoracic to the first lumbar vertebral level (except for one case in which the greater splanchnic receives a contribution from the ninth vertebral level) and are joined in a completely inconsistent manner by interconnection, com- bination or division between the splanchnic nerves. The relatively low origin of the splanchnic nerves is consistent with the condition found in animals which have not (unlike the human being) been involved in a complex pattern of evolution of the vertebral column. The lack of consistency in the manner in which the splanchnic nerves are formed might be considered to be indicative of a rather primitive (or unstabilized) developmental pattern in this part of the sympathetic nervous system; this constitutes, however, only a speculative possibility, since considerable variation in the pattern of formation of the splanchnic nerves does exist in other species. 540 THe UNIveRSITY SCIENCE BULLETIN Unlike the cat, dog, rhesus monkey and human being in which the caudal (inferior) mesenteric ganglion is intimately associated with the artery of the same name, the caudal mesenteric ganglion in the opossum is located, not on the caudal mesenteric artery, but in the mesentery approximately twenty millimeters above the bifur- cation of the abdominal aorta. As in the other forms mentioned, the caudal mesenteric ganglion receives fibers mainly from the in- termediate mesenteric plexus, the celiac ganglionic complex and the splanchnic fibers of the lumbar ganglia, and it serves as the main source of the sympathetic fibers to the descending colon, rectum and gonads. The position of the caudal mesenteric gan- glion appears to be related to the position of the caudal mesenteric artery; it would seem that the ganglion has undergone partial migra- tion from its usual position toward the unusually placed artery. This suggests that the pattern of placement of sympathetic ganglia and fibers in close association with blood vessels may well be of more significance than is commonly assumed. In the sacral region, it is of interest that there is no ganglion impar recognizable in the opossum; it is probable that the communi- cations between the two sympathetic trunks may have isolated gan- glion cells in association with them, and structurally and functionally replace the true ganglion of other forms. The sacral autonomic nerves or pelvic nerves are derived only from the first and second sacral nerves in the opossum; the equivalent nerves arise from the second and third sacral spinal nerves in cat (Reighard & Jennings, 1961), from second, third and fourth sacral spinal nerves in the human being (Kuntz, 1953) and from the seventh lumbar, first and second sacral spinal nerves in the rhesus monkey (Kuntz, 1933). The variation in the levels of origin of the pelvic nerves in the forms mentioned are, however, probably related to variations in number- ing of vertebrae rather than to changes within the central nervous system and its branches. The number of sources contributing to the pelvic viscera (es- pecially the gonads) in the opossum is similar to that in the human being; the gonads in both receive fibers from the renal plexus by way of the ureter, from the gonadal ganglion and intermediate mesenteric plexus via the gonadal artery, from the hypogastric plexus by way of the branch of the artery of the same name and from the pelvic plexus. The rich contribution of fibers from sources situated as far cranially as the renal plexus seems to be here, as in Nervous SYSTEM OF THE OPOSSUM 541 other forms, an indication of a developmental process in which the elongation of the body and descent of the viscera extend the fibers to their final position. Among the cranial autonomic ganglia in the opossum, only the submaxillary ganglion differs significantly from that in the human being in relative size and in location. It is less than one millimeter in diameter, the smallest of the four ganglia. Since the submaxillary gland is located farther from the oral cavity in the opossum than it is in the human being, the submaxillary duct is comparatively longer, and the submaxillary ganglion is located far back on the common duct immediately before it is formed by the union of the submaxillary duct and the duct of the sublingual gland. It seems reasonable to assume that the ganglion has been carried posteriorly in the course of caudal migration of the gland in the growth process, and that its small size and diffuse nature may be indicative of synaptic stations firmly associated with the gland itself. Although fibers from the otic ganglion to the auriculotemporal nerve were not recognizable, a filament from the lateral aspect of the otic ganglion to the mandibular nerve was observed. This latter filament may be secretory in nature and may eventually follow the auriculotemporal nerve, a branch of the mandibular nerve, to reach the parotid gland. SUMMARY The general morphology of the autonomic nervous system of the opossum, as it is revealed by gross silver staining and dissection, is described in some detail. The thoracolumbar division arises from the last cervical vertebral level of the spinal cord cranially to the third or fourth lumbar vertebral level caudally, and is fundamentally similar in its morphology to that of other mammals. The cervical portions of the sympathetic trunks do not communicate with the vagi and communicate with the thoracic portions primarily by way of the ansa subclavia. The prevertebral ganglia of the abdominal region are poorly differentiated, one from another, and no definitive ganglion impar is recognizable. The four cranial autonomic ganglia are similar to those in other mammals, except that the submaxillary ganglion is small and diffuse and is located immediately anterior to the common duct formed by the union of the submaxillary duct with the duct of the sublingual gland. 542 THe UNIVERSITY SCIENCE BULLETIN LITERATURE CITED BaBKIN, B. P. 1949. Antagonistic and synergistic phenomena in the autonomic nervous system. Trans. Roy. Soc. Can., vol. 40, Sect. V, pp. 1-25. BERNHEIM, R. 1934. Action of drugs on the isolated intestine of certain teleost fish. Jour. Boeke, J. 1949. The sympathetic endoformation, its synaptology, the intestial cells, the periterminal network and its bearing on the neuron theory— discussion and critique. Acta Anat., vol. 8, pp. 18-61. BroEk, A. J. P. v. D. 1908. Untersuchungen iiber den Bau des Sympathischen Nervensystems der Siugetiere. Morph. Jahrb. (Gegenbaur), vol. 38, pp. 532-541; 574-588. Buttock, T. H. 1945. The anatomical organization of the nervous system of Enteropneusta. Quart. Jour. Micro. Sci., vol. 86, pp. 55-111. VAN CAMPENHOUT, R. 1930. Historical survey of the development of the sympathetic nervous system. Quart. Rev. Biol., vol. 5, pp. 217-234. CANNON, W. B. 1936. Bodily Changes in Pain, Hunger, Fear and Rage. 2nd ed., New York: Appleton, pp. xvi + 404. CHRISTENSEN, R. 1936. Sympathetic and parasympathetic nerves in the orbit of the cat. Jour. Anat. vol. 70, pp. 225-233. DANIEL, J. F. 1943. The Elasmobranch Fishes. Berkeley: University of California Press, pp. 258. DawypoFF, C. 1948. Classe des Entéropneustes. In Traité de Zoologie 11, pp. 379: Ed. p-p Grassé. Paris: Masson. FRIEDMAN, M. H. F. 1935. A. study of the innervation of the stomach of necturus by means of drugs. Trans. Roy. Soc. Cana., vol. 29, Sect. v, pp. 175-185. GELLHORN, E. 1943. Autonomic Regulations. New York: Interscience, pp. vii + 365. Goopricu, E. S. 1958. Study on the Structure and Development of Vertebrates. London: Constable, pp. 770-785. Hinsey, J. C., K. Hare and G. A. WoLF 1942. Structure of the cervical sympathetic chain in the cat. Anat. Rec., vol. 82, pp. 175-179. Nervous SYSTEM OF THE OPossuM 543 Hirer, A. 1934. Sympathisches Nervensystem und Nebenniere. In Handbuch der Vergleichenden Anatomie der Virebeltiere, vol. 2, pp. 685-687; 735- 776. Ed. Bolk et al. Berlin und Wien: Urban und Schwarsenberg. KRAUSE, R. 1923. Mikroskopische Anatomie der Wirebeltiere. Berlin & Leipzig: Walter de Grugter, pp. 88-91. Kuntz, A. 1911b. The evolution of the sympathetic nervous system in vertebrates. Jour. Comp. Neur., vol. 21, pp. 215-236. 1933. The Autonomic Nervous System. In Anatomy of the Rhesus Mon- key. Ed. C. C. Hartman and W. L. Straus, Jr., pp. 328-338. Baltimore: Williams and Wilkins. 1951. Visceral Innervation and its Relation to Personality. pp. 115-135. Springfield: Thomas. 1953. The Autonomic Nervous System. 4th ed., pp. 15-605. Philadel- phia: Lea and Febiger. 1956. Components of splanchnic and intermesenteric nerves. Jour. Comp. Neuro., vol. 105, pp. 251-268. 1958. Das Autonome Nervensystem. In Handbuch der Zool., Bd. VIII, 11th lieferung, 7(4), pp. 1-42. Berlin: Walter de Gruyter. Lutz, B. R., and L. C. Wyman 1932. Reflex cardiac inhibition of branchio-vascular origin in the elasmo- branch, Squalus scanthias. Biol. Bull., vol. 62, pp. 10-17. Mizeres, N. J. 1955. The anatomy of the autonomic nervous system in the dog. Amer. Jour. Anat., vol. 96-97, pp. 285-318. Mucwer, E. and G. LitjestRAND 1918. Anatomische und experimentelle Untersuchungen iiber das autonome Nervensystem der Elasmobranchier nebst Bemerkungen iiber die Darmerven bei den Amphibien und Saugetieren. Arch. f. Anat. u. Physiol. Anat. abtlg., pp. 137-172. NICHOLLS, J. V. V. 1934. Reaction of the smooth muscle of the gastrointestinal tract of the state to stimulation of autonomic nerves in the isolated nerve- muscle preparation. Jour. Physiol., vol. 83, pp. 56-67. Nicot, J. C. CoLIn 1952. Autonomic nervous system in lower chordates. Biol. Rev., vol. 27, pp. 1-49. REIGHARD, J. and H. S. JENNINGS 1961. Anatomy of the Cat. 38rd ed., iii + 432. New York: Holt, Rinehart and Winston. RICHTER, C. 1950. An ideal preparation for dissection of spinal, peripheral and auto- nomic nerves of the rat. Science, vol. 112, no. 3897, pp. 20-21. 544 Tue UNIVERSITY SCIENCE BULLETIN TERNI, 1931. Il Simpatico cervicale degli Amnioti (Ricerche di Morfologia com- parata). Sschr. Anat. u. Entwgsch., vol. 96, pp. 389-426. Wyman, L. C. and B. R. Lutz 1932. The effect of adrenalin on the blood pressure of the elasmobranch, Squalus acanthias. Biol. Bull., vol. 62, pp. 17-22. Youne, J. Z. 193la. On the autonomic nervous system of the teleostean fish, Urano- scopus scaber. Quart. Jour. Micr. Sci., vol. 74, pp. 491-535. 1931. The pupillary mechanism of the teleostean fish, Uranoscopus scaber. Proc. Roy. Soc. Lond. B., vol. 107, pp. 464-484. 1933a. The autonomic nervous system of salachians. Quart. Jour. Micr. Sci., vol. 75, pp. 571-624. KEY TO ABBREVIATIONS A—aorta AG—adrenal gland AP—aortic plexus ARG—aorticorenal gang. AS—ansa subclavia ATN—auriculotemporal n. AbN—Abducens n. CCA—common carotid a. CCSG—cranial cervical sym. gang. CG—ciliary gang. CLN—cranial laryngeal n. CMG—cranial mesenteric gang. CN—cardiac n. CP—carotid plexus CR—communicating ramus CST—cervical sym. trunk CTN—chorda tympani n. CaCSG—caudal cervical sym. gang. CaMG—caudal mesenteric gang. CeG—celiac gang. CoP—coronary plexus DA—descending aorta DCP—dorsal cardiac plexus DD—ductus deferens DPN—deep petrosal n. E—esophagus EN—ethmoidal n. EP—esophageal plexus EPM—ext. pterygoid m. GA—gonadal a. GAA-~ gastric a. GG—gonadal plexus GSN—greater splanchnic n. HP—hypogastric plexus HeP—hemmorrhoidal plexus IAN—inf. alveolar n. IMP—intermediate mesenteric plexus IN—infraorbital n. IOM—inf. oblique m. IPM—int. pterygoid m. LG—lacrimal gland LI—large intestine LN—lingual n. LR—long root LSG—lumbar sym. gang. LSN—lesser splanchnic n. LSPN—lesser superficial petrosal n. LSpN—lumbar splanchnic n. LsSN—least splanchnic n. MCG—nmid. cervical gang. MN—mazxillary n. MPN—maior palatine n. NG—nodose gang. OG—otic gang. ON—oculomotor n. ON-Id—oculomotor n., inf. div. ON-Sd—oculomotor n., sup. div. OP—ovarian plexus OpN—optic n. PA—phenic a. PCN—n. of pterygoid canal PG—parotid gland PNN—post. nasal n. PP—pulmonary plexus PeG—pelvic gang. PeN—pelvic n. Nervous SYSTEM OF THE OpossuM PeP—pelvic plexus PrG—prostate gland PuN—pudendal n. R—rectum RN—recurrent n. RP—renal plexus SA—subclavian a. SCN—short ciliary n. SG—stellate gang. SLD—sublingual duct SLG—sublingual gland SMD—submaxillary duct SMGl—submaxillary gland SP—splenic plexus SPG—sphenopalatine gang. SPLM—sup. palpebral levator im. SPN—superficial petrosal n. SRM—superior rectus m. T. trachea Te—testis TN—trochlear n. TSG—thoracic sym. gang. U—uterus UB—urinary bladder UP—ureteric plexus UgA—urogenital a. UgS—urogenital sinus Ur—ureter UrP—uteric plexus Vi—ophthalmic n. V2—maxillary n. V3—mandibular n. V3-ab—mandibular n., ant. br. VCP—ventral cardiac plexus VN—vagus n. VP—vaginal plexus VeN—vertebral n. VeP—vesical plexus 545 546 THE UNIVERSITY SCIENCE BULLETIN Ficure 1 Fic. 1. The thoracic portions of the sympathetic trunks and vagus nerves, and the aortic and esophageal plexuses in Didelphis marsupialis. 1 547 548 ‘il witnar (lt! Tue UNIVERSITY SCIENCE BULLETIN Ficures 2-4 NG UME Q ery eA ak \ 4 LtRN CN CCA Lt VN LtMCG oe oe CLN(Lt) i} LtCCSG Fic. 2. The cervical sym- pathetic trunks and vagus nerves, ventral view. Note relationships to major struc- tures of the neck and to car- diac plexus. x 1.25 Nervous SYSTEM OF THE OPpossuM 549 Clr CP LeEeRE RtCCSG RtCLN LtVN RtRN INIA - RtMCG LtMCG - Rt VN LtCaCSG LtSA RtCCA RtCaCSG LIAS th, § RtAS RtSA LtVeN Rt VeN VCP CNN RtRN LtSG RtSG LtRN RtPP LtPP DCP RtCoP CoP TST Fic. 8. Left cervical sym- Fic. 4. Right cervical pathetic trunk, left vagus sympathetic trunk, right va- nerve and their contribution gus nerve and their contri- to cardiac plexus; ventral butions to cardiac plexus; view. x 1.25 ventral view. 1.25 22—1367 550 Tue UNIVERSITY SCIENCE BULLETIN FicurE 5 Fic. 5. The lumbar sympathetic trunks and adjacent pre-aortic ganglia and plexuses; ventral view. x 1 NERVOUS SYSTEM OF THE OPOSSUM Ficure 5 / See dol THe UNIVERSITY SCIENCE BULLETIN UB VeP PuN PeG 6 Fic. 6. Pelvic plexus in male specimen. Pelvic viscera have been turned toward the right, and the left half of the plexus is represented. Nervous SYSTEM OF THE OPpossuM 553 FIGuRE 7 UB | He = mee a Me VeP Fic. 7. Pelvic plexus in female specimen. View is as in fig.6. x 1. 554 THE UNIVERSITY SCIENCE BULLETIN Ficure 8-9 SPLM&SRM == EGLO cee Fic 8. The ciliary ganglion and its relationships to major struc- tures of the orbital region; semidiagrammatic. x 1.12 Fic. 9. The spenopalatine ganglion. » 1.12 Nervous SYSTEM OF THE OPpossuM 555 Ficure 10-11 1/1 DES Fic. 10. The submaxillary ganglion and its significant relation- ships; diagrammatic. x 1.12 Fic. 11. The otic ganglion and its major connections; diagram- matic. x 1.12 556 THE UNIVERSITY SCIENCE BULLETIN Eyes Lacrimal gland, nose and palate Submaxillary gland - Sublingual gland Parotid gland Blood vessels in head region Heart Larynx, trachea, bronchi, lungs Na SSN ESS z 2 Up / om / 7 4 / . ° . . ° \ ~e.e.. LSophagus Stomach and spleen i eceoee Liver \ N (NaN eee. Pancreas é ~S Adrenal gland and No NOs kidneys N ps ..ss. Small intestine N XN Lor eoeee Large intestine Cy = soos. Rectum ill ‘ Ee) 7 4 iS SN .eee. Bladder and ureters ~ © == => S wees. “onads Fic. 12. Schematic summary of the distribution of the autonomic nervous sys- tem in Didelphis marsupialis. X< 1 THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Voi. XLV] JUNE 7, 1965 [No. 7 The Structure and Innervation of the Venom Glands in the Tail of the Salamanders (Ambystoma) BY Grant A. Mason, Jr. James L. HA PauL GIBBONS ROOFE Department of Anatomy, University of Kansas, Lawrence, Kansas ABSTRACT 1. Twenty-two adult salamanders were used in this experiment, three for histological examinations, five for autonomic drug experiments, and the re- mainder for stimulation of the sectioned spinal cord. 2. The histological examination revealed nerves running among the smooth muscles of the acinus walls. 8. The autonomic drug experiments indicated that the nerves were from the sympathetic nervous system. 4, Stimulation of the sectioned spinal cord disclosed that the different levels of the thoracic and lumbar portions of the spinal cord (segments 10-15), when individually excited for a short period of time, would cause a segmental secretion on the tail. 5. Prolonged excitation of segments 10-15 will eventually cause secretion to appear on the entire dorsal surface of the tail, indicating the presence of a nerve syncitium. 6. This entire study proved that the walls of the poison skin glands in the tail of the salamander are innervated by sympathetic nerve fibers that have their cells of origin in spinal cord segments 10-15. INTRODUCTION The existence of “poison” skin glands in the tail of the salamander has been known for a great many years. According to Francis (1934), Leydig (1876), Pfitzner (1880), and Brasch (1894) were among the earliest known investigators to do research in regard to the development and histology of the “poison” glands. Zalesky (1866), also cited by Francis, was the first worker in the field to elucidate the chemical nature of the substance secreted by the (557) 558 THE UNIVERSITY SCIENCE BULLETIN venom glands, indicating that there probably had been some work done on these skin glands and their secretion prior to his investiga- tions. The above mentioned authors, along with more recent studies, Noble (1931), point out that the skin glands are of two kinds, mucous and poisonous. However, an extensive review of the literature does not reveal the mechanisms involved in the activation of the glands and the exact manner of their innervation. A limited number of papers have been published on the autonomic nervous system of the sala- mander, Francis attributes to Andersson (1892), and Jaquet (1900), the first investigations in this area, the latter one being based upon the former and thereby suffering from some incom- pleteness found in the first work. Hoffmann’s (1902) paper on the development of the sympathetic nervous system (sympathetic here referring to the autonomic nervous system as a whole) added sev- eral important details in this field. The literature, since the early 1900’s is lacking in references to these venom glands and their innervation. This paper deals with the relationship between the glandular structure and the autonomic nervous system. This study was undertaken in order to determine the true nature, morphology, and innervation of the “poison” glands in the tail of the salamander. For this project the authors had available a limited number of adult salamanders, Ambystoma tigrinum melanostictum (Tiger Salamander) collected from the Jackson Hole area in Wyoming, and a large quantity of adult Ambystoma annulatum obtained through a biological supply house. Histological examina- tion of the skin glands, gross dissection of the spinal cord, and injection of autonomic drugs, coupled with electrical stimulation, were made. The results indicate that the venom glands in the tail of the salamander are under the control of the sympathetic nervous system. SURVEY OF THE LITERATURE There is a scarcity of literature related to the “poison” skin glands and their direct innervation. Therefore this survey was conducted in two phases: one, that covering the glands of the skin, and two, that covering the autonomic nervous system of the salamander and that system’s relationship to the innervation of the skin glands. The literature dealing with the skin glands specifically is quite extensive. After the beginning of the 1900’s there is a definite lack of pertinent references. The most recent reference that the VENOM GLANDS OF THE SALAMANDERS 559 authors found was that of Francis (1934). This monograph, in regards to the skin glands and the autonomic nervous system, presents brief summaries of the more pertinent details found in many of the earlier works. The earliest report of these “poison” glands is by Ascherson (1840). Esterly (1902) says that Ascherson was the first to make the distinction of the two types of skin glands in an investigation of the glands of the live frog. It was Engleman (1872), however, who first used the terms “Schleimdrusen’ and ‘Kornerdrusen, thus recognizing the existence of the two types of glands. The literal translations of “Schleimdrusen’ and ‘Korner- drusen’ are slime-forming gland and granular gland. This would seem to indicate that Englemann was using the word ‘Schleim- drusen’ to refer to mucous glands and the word ‘Kornerdrusen’ to signify venom or “poisonous” glands (serous glands). Most in- vestigators share the opinion that there are two morphologically different types of glands yet the criteria for the differentiation has not always been the same. The early classification such as non-contractile versus contractile, clear versus granular, or mucous versus poison were based on the difference in structure, physiological activity, and secretion. Daw- son (1920) says that Englemann’s 1872 classification is accepted to a degree. However, there is a diversity of opinion regarding the granular and poisonous glands. Dawson points out that Bruno (1904) classified the glands in Rana esculenta as holocrine and merocrine glands. It is not clear just what Bruno meant by this classification. Bruno probably was referring to the poisonous glands as being the holocrine type and the mucous glands as being the merocrine type. Physiological studies before and after Bruno as well as experimentation by Phisalix (1900), Hubbard (1903), and Shipley and Wislocki (1915) have proven conclusively that the skin of many amphibians secretes a poisonous substance. Accord- ing to Phisalix, cited by Dawson, the poisonous glands are found not only on the dorsal surface, as most investigators had stated, but are also located on the ventral surface of the salamander. Previously, the majority of investigators who described the location and distri- bution of the poison glands stated that they are found in the dorsal skin only. In this case, dorsal refers to an area from the mid-lateral line of one side of the tail, over the dorsal aspect of the tail of the salamander, to the mid-lateral line on the other side of the tail. Several investigators have supported the concept of one type of skin gland. The skin of Proteus, according to Bugnion (1873), 560. THe UNIVERSITY SCIENCE BULLETIN as cited by Dawson, has only one type of gland corresponding to the mucous type of Dawson. Bugnion’s results were supported by Muhse (1909) and Phisalix (1912). The Muhse paper established this idea on a firm basis. She believes that there exists in the cutis of the toad, only one kind of gland. She goes on to say that the more recent classification into mucous or poison, nuclear or granular, is based on a difference in the epithelial structure and in the secre- tion produced. This view had also been supported by Englemann (1872), Seeck (1891), and Bristol and Bartelmez (1908). Several investigators, Calmels (1883), and Junius (1896), according to Muhse, have given evidence or expressed their belief in one kind of cutaneous gland. He has excluded from his discussion, the glands of the ventral surface, which are mucous. He makes the statement that the poison glands which occur only on the dorsal surface differ from the mucous glands of the ventral side, in that they contain a milky secretion produced by the poison cells of these glands. He makes no mention of smooth muscle fibers about the glands. It is probable from his description that he has seen the muscle fibers but has mistaken them for endothelial cells, which they resemble very closely. All of the authors who described one type of gland believed that the mature form was the granular—poisonous type while the mucous gland should be regarded as an immature or younger stage of the poisonous gland. Calmels (1883) and Juninus (1896), men- tioned above, are ardent advocates of this idea. The criteria for such a differentiation is usually morphological but Esterly (1902), who speaks of poison and mucous glands, uses the difference in the secretion as a basis for classification. In his work on Plethodon, Esterly says that the largest poison glands are situated on the back of the tail, lying in that portion of the skin covering the dorsal half of the tail. Esterly points out that the two types of glands are further distinguished by other features, chief of which is the staining reactions for mucous secretion, as previously reported by Nicogla (1893), and Hoyer (1890). A firmer basis for differentiation can be afforded by a physio- logical classification. Shipley and Wislocki believe “that this work (physiological) has given, . . . added evidence for the separa- tion of the cutaneous glands . . . into mucous and poison varieties.” They say “the criteria for differentiation of the glands are usually morphological” but, as in. the case of Briston and Bartelmez (1908), and Esterly (1902), who speak of the poisonous VENOM GLANDS OF THE SALAMANDERS 561 and mucous glands, the difference in secretion has been used as a basis for classification. Physiological activity, assuredly, affords a firmer, more secure basis for classifying anatomical structures than a morphologic peculiarity, which often results from the technique used in handling the material studied. Shipley and Wislocki be- lieved that their analysis had given added evidence for the separa- tion of the cutaneous glands of at least one, possible two species of toad into mucous and poison varieties. While the literature on the poison skin glands of various Batra- chians is fairly extensive, Bristol and Bartelmez are apparently the only authors who have more recently dealt with the toad. Bristol and Bartelmez said in a short note in Science which deals with the toad but a great deal of which holds true for the salamander: The poison glands are found only on the upper surface of the body while the mucous glands are found all over the skin . . . They are much larger than the mucous glands and extend deep down into the compact corium layers. They are surrounded by a thin layer of loose connective tissue, which contains nerve fibers and a dense network of capillaries. There is almost a continuous layer of smooth muscle fibers about the gland. Hubbard (1903) says that a macroscopic and microscopic ex- amination reveals the anatomical nature of the swellings surround- ing the glandular openings on the skin. However, he fails to follow this statement up and does not tell us what the anatomical nature is. He states that the dorsal half of the epidermis of this organ is covered with minute and thickly crowded pores which can be seen with the naked eye. Muhse indicates that the epidermis of the toad has so-called warts in the dorsal skin caused by the grouping of the glands. According to Noble, the glands are often of large size and clustered in pads as in the parotid glands of the common toad or in ridges, as along the back of many species of Rana. Reese (1905) did considerable work on the Giant Salamander. He was not able to determine whether or not there were two types of glands present in the epidermis. He did state that there were three distinguishable regions, an epidermal layer, a fibrous layer, and a muscular layer. Perhaps the best way to sum up the controversy concerning whether one or two types of glands exist, would be to quote Dawson who said: “Two types of quite different glands are found. There is nothing to indicate that the smaller glands are young forms in the development of the granular ‘poison’ type. No genetic relationship between the two types has 562 THE UNIVERSITY SCIENCE BULLETIN been found. The glands are of two distinct types that differ in development, histological structure, in the staining reaction of their secretion, and in their physiological activities.” The preponderance of authors seem to agree on the definitive structure of the poison glands. Esterly has one of the best descrip- tions of glandular morphology. He contends that a layer of con- tractile or smooth muscle fibers is found between the connective tissue layer and the giandular epithelis. These fibers were first shown by Hensche (1856). The muscles of the large glands are arranged in a single layer and have a general moridional direction on the gland, converging toward the upper pole. The fibers are usually simple but may be branched, this occurring mostly on the lower part of the gland. The muscles do not form a continuous sheet about the gland; the individual fibers are separated by spaces of greater or lesser extent. Esterly also says that he was not able to find with certainty, muscles on the glands which were mucous in nature. The existence of a sphincter or constrictor muscle for the glands has been claimed by Schultz (1889), quoted by Esterly, who described a band of muscle fibers running around the neck of the gland. This observation, however, has been contested by Drasch (1889). Esterly also was unable to find such a structure in Plethodon. However, both dilator and constrictor muscles occur about the mouths of the poison glands of Plethodon. The fact that the constrictor and dilator fibers lie entirely within the epidermis need not contend against their having the function of muscles, for according to Esterly, it has been well established that the intrinsic gland musculature has an ectodermal origin. Reese holds the opinion that the poison glands are not surrounded by a muscular layer; that there is a fairly distinct basement mem- brane surrounding the epithelium of the gland; and the muscular layer which is sometimes described cannot be clearly seen. He dis- agrees with Drasch who reported in the salamander a complete layer of smooth muscle around the periphery of large poison glands. There is good evidence in fish, chiefly in Fundulus, according to Spaeth (1916) that the melanophere may be a disguised type of smooth muscle cell. Morphological and embryological evidence does not permit a clear differentiation between melanophores and smooth muscle cells. Spaeth’s physiological data is by far the most conclusive support of the parallelism between smooth muscle and melanophores. Muhse, as previously noted, goes into great detail and is quite explicit about the glands. Her description of the wall of the VENOM GLANDS OF THE SALAMANDERS 563 acinus is clearly stated. It is her belief that in some cases the wall of the acinus is thin and the muscle fibers form a more or less continuous sheath. Muhse says that Drasch is the only investiga- tor who has made any reference to a substance which encloses the muscles fibers and on which rests the epithelia. The muscle fibers are elongated and spindle shaped. Several fibers are re- quired to complete the circuit of the gland. There is, therefore, no definite arrangement of the nuclei about any given part of the acinus. Seeck (1891) describes similar structures, i. e. the spindle cells, but considers them replacement cells for the epithelia. He denies that they are muscles cells. The innervation of the smooth muscles related to the poison glands is not well understood. Noble makes the general statement that the autonomic nervous system consists of a cranio-sacral or parasympathetic outflow and a thoraco-lumbar or sympathetic out- flow. The extent to which the complex sympathetic responses of defense are organized in the amphibian as compared to the mam- mal is not clear. There is little reason to doubt that the bodily defenses in the amphibian are mobilized in accordance with our basic concepts of the autonomic nervous system. It was pointed out by Francis that there are only three investiga- tions published on the complete autonomic nervous system of the salamander, Andersson’s, Jaquet’s, and Hoffmann’s. Andersson’s arrangement is the most convenient to describe the system in three divisions: (a) cephalic; (b) cervical and abdominal; (c) caudal. The caudal portion, in which the present paper is primarily in- terested, consists of a double chain lying along the caudal artery. The ganglia are approximately segmental in arrangement. Connec- tions between the two sides are fairly frequent. The histological detail of the innervation of the glands was de- scribed by Esterly. Among the workers who have been interested in this problem are Canini (1883), Frenkel (1886), and Massie (1894). The innervation of the glands received less attention. Eckhard (1849) first demonstrated that the glands could be emptied by stimulating the anterior roots of the cerebrospinal nerves, but did not consider the structure of the nerve endings. Eberth (1869) reported that there was a network of very fine fibers close upon the glands and Openschowski (1882) described a network of nerves surrounding the glands as well as an intercellular net. Drasch (1889) also experimentally proved by nerve stimulation secretion can be obtained from the glands, as did Phisalix-Picot (1900). Loeb (1896) has also shown how closely the innervation of the 564 THe UNIVERSITY SCIENCE BULLETIN glands of Ambystoma are connected with the central nervous system. In 1898 Herrick and Coghill were able to show the exist- ence of an intimate connection of nerve fibers with the walls of the glands, but were unable to discover the exact relationship of the fibers to the gland cells. They also described the plexus of nerves beneath the corium as being composed of two types of fibers; larger ones connected with the nerve bundles of the central nervous system, and smaller ones originating in ganglion cells in the corium. Schu- berg (1903) has conflicting views, contending that many or all the nerve bundles described are really connective tissue bundles, and that the ganglion cells are the “Mastzellen” which he himself de- scribes. Concerning the work of Herrick and Coghill, Massie supports the arrangement of the fibers beneath the corium and also shows that the nerves end on the muscles of the “ental” glands. “Ental” is a word taken from Herrick and Coghill’s work and connotes the poison type of gland. These authors found that the nerve fibers passing from the nerve bundle plexus under the corium are in- timately connected with the ental glands, and seem distinct from the nerves supplying the muscles. The endings upon the smooth muscles are shown both by Cajal’s silver method and Mallory’s fuchsin stain, and in some cases are typical as described by Huber and DeWitt (1897) and Coghill (1899). In many cases fine branching fibers can clearly by seen lying upon the muscle layer. Two types of nerve endings have been described. Huber and DeWitt report that the fibers ultimately become delicate, slender twigs which, without terminal expansions, always lie on a muscle fiber and end there. Herrick and Coghill have contested that the muscles are supplied by nerves with typical endings of expansions or bulbs, as well as by fine twigs without terminal expansions. Their investigations also revealed the possibility of a connection between the nerves enveloping the glands and the gland cells, but they were not able to demonstrate it. Both the musculature and the epithelia of the granular glands have a direct nerve supply. The gland cells are surrounded by a basket work of fibers, which in some cases have terminal expansions lying on the nuclei. Spaeth was previously cited as stating that the melanophores were possibly disguised smooth muscle cells. Concerning the innervation of melanophores, Spaeth has this to report; “as we know, the activity of vertebrate smooth muscle is normally con- trolled through fibers of the sympathetic nervous system. Voluntary VENOM GLANDS OF THE SALAMANDERS 565 motor connections do not ordinarily occur.” The innervation of the melanophores has been satisfactorily demonstrated histologically by Ballowitz (1893) and physiologically by Pouchet (1876) in several species of Teleosts. Satisfactory histological demonstra- tions of the nerve endings in the melanophores of amphibians and reptiles have not been recorded. Simmerman (1878) and Hooker (1912) have demonstrated physiologically the sympathetic innerva- tion in the melanophores of the frog and Carlton (1904) has made similar observations in Anolis. Thus, the question of melanophores possibly being disguised smooth muscle cells seems to be unsettled. Herrick and Coghill state that they succeeded in securing ex- cellent preparations in the toad in which they find it easy to trace non-medullated fibers from the ectad plexus of the corium into the most intimate connection with the superficial walls of the glands. These glands are very large and quite active in the toad. The fibers are of small caliber, excessively numerous, and envelop the whole gland in what at first looked, to the two investigators, like a closely woven reticulum, but a closer study showed that the ap- pearance was due to the repeated dichotomous branching of a large number of distinct nerve fibers. These fibers cross at slightly different levels and there was little doubt, in most cases, of the complete distinctness of the fibers as they crossed. The pyridine-silver nitrate-pyrogallic acid method of Ranson gave Dawson the best results in staining nerve fibers. This method did not affect the sheaths but left the axis cylinder black. With this procedure, Dawson was able to observe nerves forming a wide mesh plexus in the subcutaneous connective tissue. Small bundles passed out from these plexi to the smooth muscle of the glands. He found, however, no evidence of the presence of a stratum of ganglion cells as did Herrick and Coghill, or Coghill himself. Dawson also disagrees with Esterly who found fibers which ran for a long distance under the dermis and eventually turned upward toward the epidermis. It was Dawson’s belief that only free inter- cellular fibers were found in the epidermis. The fact that these fibers seldom branched and were never found to terminate in small knobs or plates has been described for many Amphibia. The dis- tribution of the nerves to the glands was not apparent to Dawson because of their extreme fineness. Nor did he observe perinuclear baskets of nerve fibers, such as Esterly has described for the poison glands of Plethodon. 566 THe UNIVERSITY SCIENCE BULLETIN As mentioned previously in referring to Herrick and Coghill’s paper, the glands in the skin of the frog are of two classes, dis- tinguished as the ental and ectal series. The ental series refers to the poison glands while the ectal series designates the mucous glands. The glands of both series are distributed regularly through- out the skin of the head. The peripheral cells of the ectal series are approximately cubical, becoming lengthened or depressed ac- cording to the shape of the gland and their position in it. The walls of these glands are made up of such cells, supported externally by the fibers of connective tissue. There appears to be no differ- entiated muscular elements connected with these cells. The glands of the ental series have a peripheral structure that is entirely dif- ferent. As with the ectal series, there is a tendency for the con- nective tissue to fold about them, but within this tissue, lying compactly on the surface of the gland, is a complete tunic of non- striated muscle cells. On close examination it was discovered that there were a number of tunic cells sufficiently impregnated to show conclusively that the fibers within their varicosities lie, not between the cells, but upon them. Herrick and Coghill were unable to trace these fibers to an entrance into a bundle or as a single fiber for any distance beneath the corium, as was done in the case of the terminal fibers. The function of the fibers arising from the lower stratum, seems to be more intimately related to the glands of the ental series, for these fibers pass peripherally in large num- bers and embrace the entire surface of each gland. These fibers do not seem to be identical with those described as innervating the tunic cells of the ental series. It seems that there are two groups of nerves passing to the glands of the ental series; one attaching by typical endings to the enveloping muscle cells, and the other ramifying profusely over the surface of the gland. The nerve fibers were first described in 1862 by Kolliker as cited by Huber and DeWitt (1897). It is now believed that the endings on the involuntary muscle tissue are the neuroaxes of sympathetic neurons, situated in, or at some more remote point from the smooth muscle in which such endings are found. Two very contradictory views have been expressed on the ul- timate ending of these nerves in the involuntary muscle tissue. On the other hand, it is asserted positively that the nerve fibers terminate on the muscle cells, while on the other hand, some in- vestigators find that the nerve fibers terminate in the muscle cell, on or in the nucleus. The results of Huber and DeWitt confirm VENOM GLANDS OF THE SALAMANDERS 567 the observations of those writers who contend that the free ending of the terminal fibers end on the smooth muscle cells. According to Hubbard (1903), some doubt has been cast on the theories of protection afforded these glands in the animal kingdom that have been in favor for half a century. The glands of Plethodon oregonensis secrete a milky fluid when the animal is stimulated by an induction current, either on the back or on the tail. In a similar manner the glands respond to the touch of a drop of acid or to irritation in the form of stroking with a knife blade. The secretion turns blue litmus paper red and is therefore thought to be acidic. It does not appear to be a mucoid as it is insoluble in water. Diemyctylus toresus, when stimulated electrically, also yields a copious milky secretion, not merely upon the tail, but generally over the whole dorsal surface. Batrachoseps attenuatus, on the other hand, yielded very little secretion when stimulated in any like manner (Cope, 1889). From a comparison of the structure and the action of the glands of Plethodon with those of other Batrachia in which the nature of the secretion is known, Hubbard attributes to the secretion a poisonous and protective function. He judges that the tail glands in Plethodon offer a partial protection to the animal. They may perhaps by some offensive odor or by some irritating, volatile product, ward off an enemy. Noble (1931) postulates that the poison glands protect their owner from being devoured by many possible enemies. But these defensive properties do not always protect toads and salamanders from being eaten by snakes and other Amphibia. Three alkaloids have been extracted from the poison; Samandrin—C,,H,,N.,.— being the principle one and affecting the respiratory area of the central nervous system of the predator. Since the nature of the secretion and the function of the glands is not the primary objective of this paper, little more will be said about these two subjects. The purpose of the present investigation is to firmly establish the source of innervation of these venom glands and to determine the manner in which the nerve fibers terminated in relation to these glands. MATERIALS AND METHODS Nineteen live adult salamander, five Ambystoma tigrinum mel- anostictum, and fourteen Ambystoma annulatum were used with three previously formalin fixed adult species of melanostictum to conduct this study. All melanostictum were obtained by collection 568 Tue UNIVERSITY SCIENCE BULLETIN in the Jackson Hole area of Wyoming. After collection, these animals were transported to Lawrence, Kansas, where these ex- perimental procedures were performed. The annulatum used in this study were procured through a biological supply house. Bishop’s Handbook of Salamanders (1943) lists the distribution of annulatum as “vicinity of Hot Springs, Arkansas, and Stone County, Missouri” and melanostictum as “British Columbia, Alberta, Wash- ington, Oregon, Idaho, Montana, Wyoming, North Dakota, South Dakota, and Nebraska.” All of the animals, prior to experimentation, appeared to be physiologically normal. The average weight of melanostictum was 56.1 grams and that of annulatum was 19.6 grams. The three pre- viously fixed adults, mentioned above, were sacrificed in Moran, Wyoming at the Jackson Hole Biological Research Station and also transported to Lawrence, Kansas, for further investigation. Each animal, preliminary to any investigation, was injected inter- muscularly with a paralyzing drug, Intocostrin (E. R. Squibb and Sons, New York, New York). Intocostrin is an aqueous solution obtained from purified Chondodrendron tomentosum extract. The drug is believed to have an affect similar to that of a poison called curare (curari) used as an arrow poison by South American Indians. Its paralyzing action is reported to act upon the myo- neural junction. This drug was used because it afforded normal physiological results without interfering with the action of the auto- nomic nervous system. The first five animals, Ambystoma tigrinum melanostictum, were injected with 0.30 cubic centimeters of Intocostrin; approximately 0.15 cubic centimeters in both the dorsal pelvic and dorsal pectoral musculature. Fifteen minutes after this injection each animal was placed on its back. It was to establish whether or not the animal was anesthetized by pinching one of the animal’s digits, to confirm the level of the withdrawal reflex. The skin covering the ventral aspect of the pectoral girdle was stripped away. Each of the two overlapping cartilaginous portions of the pectoral girdle was re- flected laterally and held in that position with a hemostat. This operative procedure exposed the pericardium through which could be seen the beating heart. The thin pericardium was also removed, completely exposing the heart. By means of a 1.0 cubic centimeter tuberculin syringe and a 30 gauge, % inch hypodermic needle the different chemicals used in the experiment were injected into the heart. VENOM GLANDS OF THE SALAMANDERS 569 AUTONOMIC DRUGS AND STIMULATION Animal Number 1: Approximately 0.05 cubic centimeter of Adrenalin in 0.9% saline solution was injected into the ventricular cavity of the heart. A thirty second resting period followed the injection. Immediately ensuing this thirty second latent period a shock from an electronic stimulator, Model 751, 115 volts AG, 50-60 was administered via insulated electrodes to the dorsal portion of the tail at the spinal cord levels of S-1 through S-3. The stimulus has a frequency—pulse per second of twenty, a duration of ten milliseconds, and a voltage of three. At all times during the stimu- lation the tail was closely observed with the naked eye for any evidence of secretion. After obtaining the results of this experiment and allowing a sufficient amount of time for the glands to recover, the following four chemicals were also injected individually, in the same manner and quantity, into the heart. The chemicals were Prostigmine and Acetylcholine, Atropine, and Hexamethonium bromide. These were introduced individually and followed by stimulation. The chemicals, including Adrenalin, were used in a dilution of one part to one thousand parts of distilled water. The dosage was 1.0 milliliter per kilogram of body weight. A stimula- tion of the same intensity and frequency was applied in a similar manner and location in the case of each drug. A resting period of fifteen minutes was permitted after each phase (different chemical ) of the experiment. This animal was sacrificed to be used in a staining technique for peripheral nerves by Williams (1943) whose report stated that he had obtained excellent results with amphibians. The procedure involved the macerating of tissue in KCN and immersing in Ehrlich’s Hematoxylin. Animals Numbers 2-5: The procedures of anesthetization, ex- posure of the heart, injection of the chemicals, stimulation, and observation used on animals number 2 through number 5 were identical to those used on animal number 1. The mode of pro- cedure varied, however, in the number of chemicals used. Animals number 2 through number 5 had individual injections of Atropine, Adrenalin, Hexamethonium bromide, Prostigmine and Acetylcholine as well as another chemical, Ergotoxine ethane sulfonate. After the injection of Ergotoxine ethane sulfonate and normal stimulation, the voltage was increased to seven and the frequency raised to seventy-five and the animal was once again stimulated. 570 Tue UNIVERSITY SCIENCE BULLETIN STIMULATION OF SECTIONED SPINAL CORD Animals Numbers 6-12: This group of seven Ambystoma annula- tum could not be considered to be a definitive experiment in them- selves. They did, however, lay the background for the seven following animals and the experimentation performed on them. Each of the animals in this group was used to perfect a technique or techniques that were to be used later. The dorsal musculature covering the vertebral column was removed and the vertebra were opened to expose the spinal cord. The objective was to ascertain what sections of the spinal cord would elicit a secretion when stimulated. The cord was sectioned at various levels. The tail skin was observed when a stimulus was applied above and below the level of sectioning. Animals Numbers 13-19: A similar procedure, as will be men- tioned below, was used in this step of the investigation for each of the seven Ambystoma annulatum. All animals in this group received an injection of Intocostrin in the dorsal pelvic and dorsal pectoral musculature according to their weights, i.e., 0.3 cubic centimeters per fifty grams of body weight. Under curare, the skin on the dorsal side of the thorax, half an inch on either side of the spinal column, from thoracic vertebra 1 to thoracic vertebra 15 was removed. It has been reported that the number of thoracic vertebra are not constant and can vary from thirteen to fifteen. Our study indicates only fifteen. The underlying thin spinal muscula- ture was also removed. This step exposed the bony spinal vertebra. The spines and the dorsal half of the body of each vertebra were carefully dissected away, leaving the bare spinal cord lying in a trough created by the ventral portion of the thoracic vertebra. The final step in this phase of the investigation was to individually sever each segment and stimulate it above the point of transection, observing grossly the reactions of the glands in the tail. Between severing and stimulation there was an interval of time to allow for recovery from the shock of the operation. The average time of the recovery period was fifteen minutes. The voltage of the stimulus was three, the frequency was twenty, and the duration was ten milliseconds. Each cord segment received a shock that had a length of two seconds. A twenty minute resting state followed each of the segmental stimulations. Then, each segment that caused an initial secretion on the tail was again stimulated. The current was maintained on these seg- ments until there was secretion over the entire tail or until it was evident that no matter the length or strength of the stimulus, no VENOM GLANDS OF THE SALAMANDERS Saal secretion would be brought forth. This procedure gave a great deal more insight into the nerve distribution in the tail of the salamander. This procedure was adopted as a step in the in- vestigation for the reason that a short period of stimulation on some of the caudal segments (12, 13, 14, and 15) of the thoracic cord would not cause secretion over the entire dorsal surface of the tail. SECTIONING AND STAINING THE TAIL Animals Numbers 20-22: The tails of these three animals, Am- bystoma tigrinum melanostictum, were used to prepare the his- tological specimens required by this investigation. Number 20: Immediately after death, this animal was placed in a 10% formalin solution where it remained until studied. When the animal was removed from the formalin its tail was severed from the body and prepared for dehydration. Among the necessary steps in the preparation was the removal of the vertebral column. The column was removed by placing an incision on the “nape” or ex- treme dorsal portion of the tail and one at the most ventral portion of the tail. These two incisions were continued toward each other, both splitting and going around the bony structure and meeting their fellow from the opposite side. Thus we had two pieces of tissue, a left and a right side of the tail. The amphibian skin was then dehydrated, embedded in paraffin, and serially sectioned at fifteen micra. Every tenth section was mounted and stained with Hematoxylin and Rosin. Number 21: The same procedure as above was used except that the vertebral column was removed in such a way that the tail be- came a hollow cylinder permitting free access of the fluids to all its parts. The skin was stained with a modified Margolis-Pickett (1956) Luxol Fast Blue—Oil Red O method. Number 22: The tail was cut into small sections, 10 mm. in length, after the entire animal had been fixed in 10% formalin. These tail sections were decalcified in a solution consisting of 10% nitric acid in 70% ethyl alcohol. The usual dehydration methods were used. Serial sections in paraffin were then impregnated with a modified Holmes silver technique. RESULTS AUTONOMIC DrRuGS AND STIMULATION Animal Number 1: Following the injection of 0.05 cubic centi- meters of Adrenalin in 0.9% saline, there was an immediate doubling Die THE UNIVERSITY SCIENCE BULLETIN of the cardiac rate to ten beats per five seconds. The electrodes were applied to the dorsal nape of the tail. When the current was turned on, the poison glands released a copius, white, sticky secretion within two seconds. The same shock was applied to the ventral regions of the tail but there was no evidence of any secretion. Prostigmine, followed shortly (thirty seconds) by Acetylcholine was injected into the heart in quantities of 0.05 cubic centimeters. The second injection caused a primary cardiac acceleration which was soon altered to an over-all deceleration of three beats every five seconds. Similar to the Adrenalin injection, there was a slight watery appearance on the ventral side of the tail. The third injection was Atropine. The procedures relating to quantity of the drug, injection, the waiting period, the stimulation, and observation were identical to those used in the Adrenalin and Prostigmine-Acetylcholine experiments. The results with this drug were very similar to those obtained when using the three perviously mentioned compounds. There was an immediate and copious secretion on the dorsal surface of the tail when it was stimulated. In this case there was no evidence of a watery secre- tion on the ventral surface. Hexamethonium bromide was the fourth and final chemical used on animal number |. There was none to very slight secretion on any portion of the tail when it was electrically stimulated. The only part of the tail that showed any evidence of secretion was the tip. Animals Numbers 2-5: The four autonomic drugs used on animal number | were also used on animals numbers 2-5. The methods of experimentation were the same and the results were also identical. See Table 1. However, a fifth drug was introduced into these four animals. The fifth chemical was Ergotoxine ethane sulfonate. It was injected in a quantity similar to that previously used on all the animals. The heart had almost an immediate slowing reaction. The rate was two beats per five seconds but each contraction was violent. In five minutes there was a distinctive blanching of the heart and the beats had slowed to one per five seconds. There was no evidence of any secretion elicited by stimulation other than on the last eighth of an inch of the tail. Even though the voltage and frequency as well as the length of stimulation were all significantly increased, no secretion was seen to come from the poison glands. VENOM GLANDS OF THE SALAMANDERS Dita STIMULATION OF SECTIONED SPINAL CorD Animals Numbers 13-19: Of the seven animals used in this step of the experiment, no secretion was ever observed when thoracic cord segments one through nine inclusive were stimulated. There was, however, a gradient found in segments ten through fifteen re- lating to the amount of secretion that came forth when the respec- tive segments were stimulated. See Table II. Segment ten, when stimulated, caused the entire dorsal half of the tail to secrete. Segment eleven, like segment ten, also caused the entire dorsal portion of the tail to secrete. Segments twelve through fifteen how- ever, caused a segmental secretion on the tail when stimulated. To clarify the above, we should note that segment twelve caused the caudal two-thirds of the tail to secrete, segment thirteen caused the caudal one-half of the tail to secrete, segment fourteen gave tise to secretion on the last one-third of the tail, and segment fifteen gave rise to secretion on only the very minutest portion of the tip of the tail. The secretion, caused by stimulation of thoracic cord segments ten through fifteen, had only a slight delay from the moment the electrodes contacted the cord until the milky white fluid was seen. It was observed that stimulation of segments ten and eleven caused a secretion to appear over the entire tail. The secretory product, however, did not appear simultaneously over the entire dorsal sur- face. Here too, there was a step-wise gradient, the cephalic portion of the tail secreting first and a wave of secretion appearing to spread down the tail. The entire sequence, from the moment of stimulation until the dorsal half of the tail was covered with the secretion, took less than two seconds. The most interesting results were found when the stimulation was continued over a period of time exceeding two seconds. Stimu- lation of segments ten and eleven was not necessary in this step for it had been found by previous experimentation that both of these levels gave secretion over the entire tail in less than two seconds. It was found that a continual stimulation of thoracic segments twelve through fifteen would eventually cause all of the poison glands on the dorsal portion of the tail to secrete. As one stimulated progressively down the cord, the longer the time lapse before the entire tail was covered with the poisonous secretion. Table II shows the length of time necessary for stimulation of the various cord segments before the secretion would appear. Seven 574 THE UNIVERSITY SCIENCE BULLETIN salamanders were used to compute the average time. Ambystoma annulatum was used to calculate the segmental times. Using ten specimens of A. tigrinum melanosticum with average weights (25 grams) a similar pattern of secretion was obtained in preliminary testing at the Jackson Hole Biological Station in the summer of 19860 (Roofe and Mason). THE SECTIONED AND STAINED TAIL Animals Numbers 20-22: A histological examination of the poison gland cells was made utilizing two different types of staining tech- niques. The Luxol Fast Blue—Oil Red O method did not give clear cut results. Therefore, this study was limited to those slides stained with hematoxylin and eosin and Holmes silver stain. There is a single layer of smooth muscle fibers forming a continuous sheath or thin wall around the body of the gland. This sheath was barely thicker than two muscle fibers and this condition was infrequently noted and found only when one fiber overlapped another. A majority of the walls of the poison glands were one cell thick. Thin nerve fiber endings were found lying in intimate contact with these smooth muscle fibers. This observation confirmed the very early work of Hensche. The normal poison gland cells appeared to contain three different types of secretion. There appears to be a gradient in the secretion’s relationship to the surface opening of the gland. There is a mass of large, ovoid, cellular bodies close to the surface opening, oc- cupying less than one-fourth of the total area of the gland. In the central two-thirds to three-fourths of the gland is another glandular mass composed of smaller spherical cells. The remainder of the secretion found within the unstimulated cell had the ap- pearance of “protein” material, having no cellular outline and resembling a homogeneous mass. All of the poison glands of the salamander, regardless of their volume, retained their ovoid or spherical shape. A thorough survey of all the slides made failed to reveal any type of glandular epithelia. Several of the more prominent authors of the late 1800's, including Esterly, spoke of this epithelium. DISCUSSION AUTONOMIC DRUGS AND STIMULATION The drugs employed in this experiment were chosen because of their single direct effect. Adrenalin potentiates adenergic nerve VENOM GLANDS OF THE SALAMANDERS 575 endings and it was felt, that by use of this drug, if the glands were under sympathetic control, they would release an even greater quantity of secretion than the normal amount. However, there was no device to quantitatively measure the amount of secretion under the normal or experimental conditions and the glands seemed to give less secretion every time they were stimulated. Therefore, the resting period was introduced. This resting period apparently enabled the poison glands to replenish their secretion. It had been found that continual stimulation would exhaust the glands so that they would have little or no secretion remaining in them after several prolonged shocks. Prostigmine and Acetylcholine, on the other hand, mimicked the action of the parasympathetic nervous system by potentiating cholinergic nerve endings. Prostigmine destroyed Cholinesterase which itself destroys Acetylcholine. Thus, when the Acetylcholine was introduced after the Cholinesterase had been destroyed, we had an increased action of the parasympathetic nervous system without disrupting the action of the sympathetic system. Once again the animal was stimulated and we had a copious secretion. This experiment in itself was not conclusive enough to prove whether the glands were under control of the sympathetic or para- sympathetic nervous system. The watery secretion on the ventral surface of the tail, mentioned in the results, was from the numerous mucous glands covering the body. A discharge of mucous can be obtained by handling or a slight mechanical stimulation. The third drug used was Atropine. This chemical blocked the cholinergic nerve endings. Therefore, in the way of negative proof, the copious secretion we obtained upon stimulation would tend to prove that the poison glands are under the control of the sym- pathetic nervous system. In other words, if there was a copious secretion when the parasympathetic nerve endings were blocked, the conclusion that the glands were under adenergic (sympathetic ) control would be supported. To be sure that the glands were under a nervous control simu- lated by the autonomic nervous system, Hexamethonium bromide was injected. This drug blocks synaptic transmission at the ganglion thus producing a total block of all autonomic nerve endings. Thus, when a stimulus was applied, there should be no evidence of any secretion. This drug was injected in order to determine whether the adrenal glands were causing the skin glands to be activated. If the adrenals were affecting the activity of the poison glands 576 THe UNIVERSITY SCIENCE BULLETIN we would expect a secretion in spite of the drug injection. When the stimulus was applied there was no evidence of any secretion, indicating two things. One, the glands were autonomically con- trolled and two, the adrenals had nothing to do with the secretion. However, as previously stated, we had only negative evidence that the poison glands were controlled by the sympathetic nervous system. It was thought, therefore, that his previous set of experi- ments should be carried one step further. A drug that blocked the adenergic nerve endings was needed. Ergotoxine ethane sul- fonate was chosen. Because of inconclusive data obtained from animal number 1, it was thought that this animal should not be included in Table I. With all five different drugs at hand, four separate tests were performed. The results of these tests have been tabulated in Table I. One can see that every test, when repeated, gave results identical to those of the previous experiment. Likewise, the four experiments that were performed upon animal number 1, when repeated on animals number 2 through number 5, gave results similar to those found in the first animal. The fifth drug in Table I is Ergotoxine ethane sulfonate. The results with this drug support the contention that the poison glands are under sympathetic nervous control. Since there was no secretion on stimulation, the most obvious con- clusion is that these glands are under autonomic nervous control and specifically, more the sympathetic nervous system. Ergotoxine had to be injected last because its affect was permanent in the dosage in which it was administered. In addition to this reason, the drug does not easily oxidize and would therefore remain in the system a long time, regardless of the dosage. The blanching of the heart, after the injection of Ergotoxine, was due to the imbalanced ac- tivity of the parasympathetic nervous system. The adenergic nerve endings, primarily the sympathetics on the coronary vessels, were blocked, causing an antagonistic constriction of these vessels and therefore blanching. There might be objection to the previous statement that upon stimulation after an injection of Ergotoxine, “there was no evidence of any secretion other than the last quarter inch of the tail.” This statement can be better understood if we realise that the circula- tion in the caudal quarter inch of the tail is quite poor. This could explain the slight secretion found when Ergotoxine had been in- troduced and the tail stimulated. Since the drug must reach the nerve endings by diffusion through the blood vessels and capillaries VENOM GLANDS OF THE SALAMANDERS 577 and their distribution in the area is poor, it stands to reason that some of the nerve endings would not receive any Ergotoxine and therefore remain quite effective. Williams (1943) method of staining peripheral nerves in cleared vertebrate tissue was attempted on animals one and two. The re- sults of this procedure were not fruitful enough to permit the con- tinuation of this method. The principal difficulty was the extreme fineness of the nerves thereby making it impossible to follow their course shortly after they left the spinal cord. STIMULATION OF THE SECTIONED SPINAL CoRD . Since it was determined that the glands of the tail were under sympathetic control, an attempt was made to determine the level of the spinal cord from which the fibers innervating these glands originated. Before this part of the experiment was begun nothing in the way of segmentation of the secretion was observed when the tail was grossly stimulated. As there was some quantity of secretion whenever thoracic seg- ments ten through fifteen were stimulated, we can say that the sympathetic nerves running to the walls of the poison glands have their cells of origin in these six thoracic cord segments. We can also state, as Table II indicates, that there is a segmental arrange- ment of these fibers with considerable overlap. This segmental arrangement and overlap have been pointed out in the results. Stimulation of the more caudal thoracic cord segments (12, 18, 14, and 15) for a period longer than two seconds caused the entire tail to be covered with secretion. This lead to some speculation. At first it was believed that the stimulatory messages were being conveyed to the brain, primarily the hypothalamus, and then re- layed back down to the poison glands. This, of course, would cause the slight delay noted when the electrodes are placed on the cord until the secretion appeared. Yet, this would appear to be difficult because most every known pathway that traversed back to the brain had been severed. The only logical conclusion that one is able to arrive at is that we have a local reflex arc and that there is a nerve syncitium that spreads throughout the entire tail. This syncitial arrangement would account for the slight delay before the tail is covered with the secretory product, as well as solve the problem that had arisen when all the pathways back to the hypo- thalamus and other autonomic regulatory areas had been severed. This syncitial net would also solve the problem of the wave of 578 THE UNIVERSITY SCIENCE BULLETIN secretion spreading up the tail when segments 12, 13, 14, and 15 were continually stimulated, rather than spreading down the tail as it would if the nervous messages were coming from the brain. THE SECTIONED AND STAINED TAIL Francis, when speaking of the autonomic nervous system, says that there are no special features to note other than the fairly fre- quent anastomoses between the two sides. He devotes consider- able space in his monograph to the venom glands and does not mention the manner in which the glands are innervated. He com- pletely ignores the thesis of glandular innervation. When investigating the histological composition of the poison glands, the concepts of Spaeth were borne in mind. He thought that there was a good possibility that the melanophores could be disguised smooth muscle cells. We could find no good histological evidence that they (the melanophores) were directly innervated by the sympathetic nervous system. The melanophores, themselves, were easily identified as were the darkly staining fibers that ran through and among the muscular walls of the glands. Yet it could not be demonstrated satisfactorily to us that the nerve fibers actually terminated on the melanophores. The fibers passed close to the pigment granules but in no cases were knobs, boutons, or spindly endings seen to terminate upon these pigment granules. Huber and DeWitt brought up the controversy of the true end- ings of the nerve fibers. They expressed the idea that two very contradictory views had been put forth; that nerve fibers may terminate in relation to the spindle shaped muscle cells or on or in the nucleus of the smooth muscle cell. We agree with Huber and DeWitt; the nerve terminals are on the muscle cells and in no case was it found that the free endings were on or in the nucleus of the muscle cell. This evidence corresponds to the work of Herrick and Coghill in 1898 who were able to show the existence of an intimate connection of the nerve fibers with the walls of the glands. Dawson and several other investigators felt that there was definite epithelial layers associated with the wall of the gland. However, very close observation did not reveal any structure that resembled an epithelial layer. The difference in this particular part of the glandular morphology is very likely due to the fact that Dawson’s investigations were on Plethodon while the present work is based on Ambystoma tigrinum melanostictum. VENOM GLANDS OF THE SALAMANDERS 579 We modified the histological techniques to better suit the tissue at hand. A thorough examination of the first preparations revealed that too much detail had either been lost or was not able to be observed. Both the staining technique, Hematoxylin and Eosin, and the method of removal of the vertebral column were modified as described in the materials and methods. The two modifications of the silver staining method mentioned in the materials and methods dealing with animal number 15 are as follows. The modification involved a shortening of several of the important steps (cutting the time in the impregnating solution from twenty-four hours to twenty-two hours and reducing the time in the reducing solution from two to five hours to one and one-half hours). This shortening seemed to stain the sections more in the manner desired, i. e., more contrast. The original time sequence made the slides too dark and very little structure in the way of nerve fibers was discernible. Our observations concerning the volume of the glands and thereby their shape (ovoid or spherical) followed the reports of Dawson, but were contrary to the ideas and writings of Drasch, Nierenstein (1908), Muhse, and Shipley and Wislocki. THrE REASON FOR AND MANNER OF THE SECRETION The skin of the salamander is moist and devoid of scales. It is highly glandular and contains two different types of glands, the mucous glands, whose function it is believed, is to keep the skin moist, and poison or venom glands whose function will be speculated upon briefly. The manner of expulsion of the poison and its true chemical nature is somewhat of a mystery. When a salamander is irritated by chloroform, the white, milky fluid is seen to exude from the venom glands. Likewise, Dawson has said that the “granular secretion is expelled only in responses to some violent stimuli either mechanical, chemical, or electrical.” He also believed that the secretion of the poison was due to the contraction of the smooth muscles found about the gland sacs. Seeck says that the expulsion is caused by underlying skeletal muscle. This idea of Seeck’s can not be held as true because the primary injection of Intecostrin blocked all of the skeletal myo-neural junctions, render- ing the skeletal muscle incapable of any response. Actually, later authors quite generally agreed that the expulsion of the secretion was due to smooth muscle contractions. Yet Calmels, like Seeck, did not attribute expulsion to the smooth muscle fibers about the 580 THe UNIVERSITY SCIENCE BULLETIN individual glands. Calmels did not describe or diagram, as such, a layer of smooth muscle fibers in the acinus wall. His statements regarding muscles are not clear, and Muhse felt that it was best not to attempt to state his position. The secretion of the mature glands comes to the surface of the skin in drops. It is white or creamy in color, has a bitter taste, and a rather strong, disagreeable odor, very similar to that of Jimsom weed (rank-smelling foliage). Inside the mature glandular cell, the secretion as mentioned before, is primarily granular in appear- ance and takes stain readily, so that it always has a decided color, thus easily distinguished from the mucous secretions which are only very slightly affected by stains. Noble reported that these granular glands produce a secretion which is usually more toxic than that of the mucous glands. It is believed that the secretion contains more than one poisonous con- stituent. Faust (1898), according to Noble, isolated Bufotalin, giv- ing it the empirical formula of C,,H,,O,,. Francis states that Zalesky isolated the chlorohydrate of one of the alkaloids of the poison, which he called Samandrin. Abel and Macht (1911), in analyzing the secretion of the tropical toad, Bufo agus, said that the glands are under the control of the central nervous system. Bristol and Bartelmez verified that the secretion had the reputa- tion of being poisonous. They reported that it acted only on the mucous membrane, producing a result similar to that of curare. CONCLUSIONS In order to better understand the poison skin glands of the salamander under normal and stimulatory conditions, histological examinations, stimulatory experiments coupled with sectioning of the spinal cord, and autonomic drug injections followed by stimu- lation, were made on the tails of twenty-two adult salamanders. We found that the walls of each of the poison skin glands was surrounded by a thin layer (one cell thick) of smooth muscle. In contrast to these walls, the mucous glands contained no smooth muscle in their walls. In close and intimate contact with the smooth muscle cells of the walls are tiny, slender nerve endings. These fibers are from the sympathetic nervous system and come from thoracic cord seg- ments ten through and including fifteen. By the use of autonomic drugs and stimulation we found that the poison glands in the dorsal portion of the tail are innervated VENOM GLANDS OF THE SALAMANDERS 581 TaBLE I.—Secretion Results CHEMICAL Animal I Animal II | Animal III | Animal ITV INGEOVONING 5 o'o'b:5'5 0 5 Copius Copius Copius Copius Adrenalin.......... Copius Copius Copius Copius Hexamethonium bromide..........| Nothing Nothing Nothing Nothing Prostigmine and Acetylcholine... . Copius Copius Copius Copius Ergotoxine ethane Sulfonate......... Nothing Nothing Nothing Nothing TABLE II. Geena Results of two second Length of stimulation cord segments stimulation of severed cord segments necessary to cause entire tail to secrete SCUOONOORWNe — Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Secretion over entire dorsal half of tail Secretion over entire dorsal half of tail Secretion over last two-thirds of tail Secretion over last one-half of tail Secretion over last one-third of tail Secretion only on the very tip of tail 1.4 seconds 1.8 seconds 2.5 seconds 2.8 seconds 3.9 seconds 4 4.1 seconds in a serial fashion. That is, the last thoracic segment of the spinal cord, when stimulated, would cause a secretion to appear on the very last portion or tip of the tail. Whereas a more anterior seg- ment, when stimulated, would cause secretion over a greater portion of the tail. 23—1367 582 THe UNIVERSITY SCIENCE BULLETIN The total innervation of the poison glands is bound up in a nerve syncitium. This accounts for the wave of secretion being able to spread either way on the tail, the delay in the appearance of secretion over the entire tail, and the manner in which the glands are innervated when all known pathways back to the brain have been severed. These glands appear to be protective in nature for the secretion is only released when the animal is severely traumatized. SUMMARY 1. Twenty-two adult salamanders were used in this experiment, three for histological examinations, five for autonomic drug ex- periments, and the remainder for stimulation of the sectioned spinal cord. 2. The histological examination revealed nerves running among the smooth muscles of the acinus walls. 3. The autonomic drug experiments indicated that the nerves were from the sympathetic nervous system. 4. Stimulation of the sectioned spinal cord disclosed that the different levels of the thoracic and lumbar portions of the spinal cord (segments 10-15), when individually excited for a short period of time, would cause a segmental secretion on the tail. 5. Prolonged excitation of segments 10-15 will eventually cause secretion to appear on the entire dorsal surface of the tail, indicat- ing the presence of a nerve syncitium. 6. This entire study proved that the walls of the poison skin glands in the tail of the salamander are innervated by sympathetic nerve fibers that have their cells of origin in spinal cord segments 10-15. VENOM GLANDS OF THE SALAMANDERS 583 LITERATURE CITED ABEL, J. J. and Macut, D. I. 1911. The poisons of the tropical toad, Bufo agua. Jour. Amer. Med. Assn., 56: 1531-1536. ANDERSSON, O. A. 1892. Zur Kenntnis des sympathischen Nervensystems der urodelen Am- phibien. Zool. Jahrb., Abt. f. Anat., 2: 184-210. ASCHERSON, C. 1840. Uber die Hautdrusen der Frosche. Arch. f. Anat. u. Physiol., 15-23. BALLowiI7z, E. 1893. Die Nervenendigungen der Pig mentzellen, ein Beitrag zur Kenntnis des zusammen hangesder Endverzweigungen der nerven mit dem Protoplasma der Zellen. Zeitschr. f. wiss. Zool., 56: 673-706. Bisuor, S. C. 1943. Handbook of Salamanders. New York: Comstock Publication Com- pany, Inc., 117-154. BOULENGER, G. A. 1891. The poisonous secretion of Batrachians. Ann. Mag. Nat. Sci., 8: 453-456. BrisTou, C. L. and BARTELMEz, G. W. 1968. The poisonous glands of Bufo agua. Science., 27: 455. Bucnion, E. 1873. Recherches sur les organes sensitifs qui se trouvent dans |’epiderme du protee et de l’axolotl. Bull. Soc. Vaud. Sci. Nat., 12: 259-316. CALMELS, G. 1883. Etude histologique des glandes a venin du crapaud et recherches sur les modifications apportees dans leur evolution par l’excitation electrique de Tanimal. Arch. de Physiol., Norm. et Path., 3: 321-362. CANINI, A. 1883. Die Endigungen der Nerven in der Haut des Froschlarvensch- wanzes. Arch. f. Anat. u. Physiol., 159-160. CARLTON, F. C. 1903. The Colour Changes in the skin of the so-called Florida Chameleon, Anolis carolinensis Cuv. Proc. Amer. Acad. Arts and Sci., 39: 259-276. CocuHiL1, G. E. 1899. Nerve termini in the skin of the common frog. Jour. Comp. Neur., 9-53-63 Corr, E. D. 1889. The Batrachia of North America. Bull. U.S. Nat. Mus., 34: 1-525. Dawson, A. B. 1920. The integument of Necturus maculosus. Jour. Morphol., 34: 487-590. Drascu, O. 1889. Beobachtungen am lebenden Drusen mit und ohne Reizung der Nerven derselben. Arch. f. Anat. u. Physiol., Physiol. Abth., 96-136. 584 THe UNIVERSITY SCIENCE BULLETIN 1894. Der Bau der Giftdrusen des gefleckten Salamanders. Arch. f. Anat. u. Physiol., Anat. Abth., 225-268. EBERTH, C, J. 1869. Untersuchungen zur normalen und pathologischen Anatomie des Froschhaut. Leipzig. EckKHARD, C. 1849. Uber den Bau der Hautdrusen der Kroten und die Abhangigkeit der Entleerung ihres Sekretes vom centralen Nervensystem. Arch. f, Anat. u. Physiol., 415-430. ENGLEMANN, T. W. 1872. Die Hautdrusen des Frosches. Eine physiologische Studie. Arch. ges. Physiol., 5: 498-537, and 6: 97-156. EsTErRLy, C. O. 1904. The structure and regeneration of the poison glands of Plethodon. Univ. Calif. Pub. Zool., 1: 227-268. Francis, E. T. B. 1934. The Anatomy of the Salamander. Oxford: Oxford at the Claren- don Press, v-304. FRENKEL, S. 1886. Nerv und Epithel am Froschlarvenschwanz. Arch. f. Anat. u. Physiol., 414-531. HENscuE, A. 1856. Uber die Drusen und glatten Muskeln in der ausserem Haut von Rana temoraria. Zeit. f. wiss. Zool., 7: 273-282. Herrick, C. L. and Cocuii1, G. E. 1898. The somatic equilibrium and the nerve endings in the skin. Jour. Comp. Neur., 8: 32-56. HOFFMANN, C. K. 1902. Zur Entwicklungsgeschite des Sympatheticus II. Die Entwick- lungsgeschite des Sympatheticus bei den Urodelen. Verh. kon. Akad. Wet., 3: 101-117. Hooker, D. 1912. The reactions of the melanophores of Rana fusca in the absence of nervous control. Ztschr. f. allg. Physiol., 14: 93-101. Hoyer, H. 1890. Uber den Nachweis des Mucins in Geweben mittels der Farbe- methode. Arch. f. mikr, Anat., 36: 310-374. Huspagrp, M. E. 1903. Correlated protective devices in some California Salamanders. Univ. Calif. Pub. Zool., 1: 157-170. Huser, G. C. and DeWitt, L. M. H. 1897. A contribution to the motor nerve-endings and on the nerve-endings in the muscle spindles. Jour. Comp. Neur., 7: 169-230. Jaquet, M. 1900. Anatomie comparee du systeme nerveux sympathique cervical dans la serie des vertebres. Bul. Soc. de Sciite., 240-404. VENOM GLANDS OF THE SALAMANDERS 585 Junius, P. 1896. Uber die Hautdrusen des Frosches. Arch. mikr. Anat., 47: 136-154. Leypic, F. 1876. Uber die allgemeine Bedeckungen der Amphibien. Arch. mikr, Anat., 12: 119-242. Logs, L. 1896. Sur Theorie des Galvantropismus. III uber die polare Erregung der Hautdrusen von Ambystoma druch constanten Strom. Pfluger’s Archiv., 65: 308-316. Marcouis, G. and Pickett, J. P. 1956. New applications of the Luxol Fast Blue myelin stain. Laboratory Investigations, 5: 469-470. Massie, J. H. 1894. Gland and nerve-endings in the skin of the tadpole. Jour. Comp. Neur., 4: 7-12. Muusez, E. F. 1909. The cutaneous glands of the common toads. Amer. Jour. Anat., 9: 321-360. Nicocuv, P. 1893. Uber die Hautdrusen der Amphibien. Zeit. f. wiss. Zool., 56: 409-487. NIERENSTEIN, E. 1908. Uber den Ursprung und die Entwicklung der Giftdrusen von Salamandra maculosa, nebst einem Beitrag zur Morphologie des Sekretes. Arch. mikr. Anat., 72: 47-140. Nose, G. K. 1931. The Biology of the Amphibia. New York: McGraw-Hill Book Company, Inc., vii-243. OpENCHOwskKI, T. 1882. Histologisches zur Innervation der Drusen. Pfluger’s Archiv., 37: 223-233. PFITZNER, W. 1880. Die Epidermis der Amphibien. Morph. Jahrb., 6: 469-526. PuHIsALix, C. 1900a. Recherches embryologiques histologiques et physiologiques sur les glandes a venin de la Salamandre terrestre. These de doctorat en medicine, Paris. 1900b. Recherches embryologiques, histologiques et physiologiques sur les glandes a venin de la Salamandre terrestre. Bull. Mus. Hist. Nat.. Paris, 6: PuisALtx, MMe. (Marie) 1912. La peau et ala secretion muqueuse chez le protee Anguillard et la Sirene lacertine. Bull. Mus. Hist. Nat., Paris, 18: Poucuet, M. G. 1876. Des changements de coloration sous Vinfluence des nerfs. Jour de l’Anat. et de la Physiol. norm. et path., 12: 1-90, 113-165. 586 THe UNIVERSITY SCIENCE BULLETIN Reese, A. M. 1905. The enteron and integument of Crytobranchus allegheniensis. Trans. Amer. Micro. Soc., 26: 109-120. Roore, P. G. and Mason, G. A. 1960. Preliminary Studies on Secretion in response to electrical stimulation Amblystoma tigrinum. Unpublished data. SCHUBERG, A. 1903. Unterschungen uber Zellverbindungen. Zeit. f. wiss. Zool., 74: 156-325. ScHULTz, P. 1889. Uber die Giftdrusen der Kroten und Salamander. Arch. f. mikr. Anat., 34: 11-57. SEECK, O. 1891. Uber die Hautdrusen eininger Amphibien. Inaug. Dissert., Dorpat. Suiptey, P. G. and Wistockt, G. B. 1915. The histology of the poison glands of Bufo agua and its bearing upon the formation of epinephrin within the glands. Contri. to Embry., Carnegie Instn. of Wash., 3: 71-82. SPAETH, R. A. 1916. Evidence proving the melanophore to be disguised type of smooth muscle cell. Jour. Exp. Zool., 20: 193-216. WitiiaMs, T. W. 1943. A technique for the gross differential staining of peripheral nerves in cleared vertebrate tissues. The Anatomical Record., 86: 189-195. ZALESKY, C. 1866. Uber die Fortpflanzung der Urodelen Amphibien. Schriff. Physik. okon. Ges., Konigsberg, Jahrb., 33: 21-26. THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vou. XLV] June 7, 1965 [No. 8 The Sympathetic and Parasympathetic Nerve Fibers in the Renal Cortex of the Cat BY Puyius D. Zwarycu and Paut GrBsons ROoFE Department of Anatomy, University of Kansas, Lawrence, Kansas Asstract: The results of the experiments with regard to the presence of sympathetic and parasympathetic nerves in the cat support the concept that both sympathetics and parasympathetics contribute to the innervation of the cat kidney. The results do not support the concept that there are sympathetic but no parasympathetic fibers in the kidney. The results also propose that nerves enter the renal cortex by following renal tubules as well as blood vessels. TABLE OF CONTENTS PAGE MimETO GUC tO Tener eae Gee ee eet Pale erg he ea oe an eet AV ae ae 587 NMaterialssandeNiethocdsers a. eich toon ey pe pas is eter an nt gpa 591 Resultswam cD iSCusSsiOMS es are ee en piled ne 595 Summanyaands©onclusionsSae niet Ser ee ee 602 He iteratunne wa Cite lee eye ie, eos Seana tin nN asta Sie wee REN 604. KeyetowAlb bre vilatsl OMS mi ase ae eee ane ear ae earrycee ear re eer ye 605 INTRODUCTION Pappenheim in 1841 presented the first description of nerve fibers in the parenchyma of the kidney as cited by Gruber (1933). Pap- penheim was able to trace nerve fibers by gross dissection down to arteries of 0.3 millimeter in diameter. Some years later Beale (1869) examined the kidneys of frogs, pigs, rabbits, newts, dogs, cats and man by prussian blue and carmine staining techniques and reported nerve filaments in relation to the arteries and tubules. Holbrook (1883) used a modified Cohnheim’s gold choloride method and described rich plexuses of nerve fibers around urinif- erous tubules. Bradford (1888-9) described nerves in relation to renal blood vessels but did not discuss the arrangement with regard to the renal (587) 588 THe UNIVERSITY SCIENCE BULLETIN tubules. Berkley (1893) considered that the intrinsic renal nerve network included nerves accompanying blood vessels ending on the adventitia of these vessels as well as between renal tubules. He observed a fine meshwork of nerve fibers around the glomeruli. Azoulay (1894-5) used Golgi and Cajal staining techniques to delineate nerve filaments accompanying and innervating the blood vessels and ending in a network among the tubules and the glomeruli. He reported nerve filaments both on and within the glomerular capsules with some filaments penetrating between capillary loops. He mentions that the cortex has a much richer innervation than does the medulla. Probably due to the non-specific staining results thus far obtained scientists began investigating the problem of renal innervation and function by physiological experiments. Among the first physiologists to be concerned with renal innervation and function was Quinby (1916) who did his experiments on dogs using a modified trans- plantation technique. Under ether anaesthesia he removed the right kidney from the body thus severing completely both its blood supply and nerve supply. He then replaced the kidney and re- established its blood supply but not its nerve supply. The opposite kidney was left untouched as a control. He noted that the im- mediate effect of the loss of nervous control over the experimental kidney was a period of overaction. After a lapse of time the ex- perimentally denervated kidney reacted in a normal manner. He stated that a kidney recently deprived of its nerve supply is one without vasomotor control therefore its blood vessels are dilated. Quinby maintained that his results gave no direct indication of the possibility of a secretory function of either the vagus or the splanchnic nerves. He suggested that chemical and hormonal action of substances contained within the circulating blood, in addition to vasomotor control within the kidney, will be found adequate to explain all variations and types of normal renal func- tions. Marshall and Kolls (1919) were concerned with nervous control and its relationship to renal function. They conducted a group of experiments which included the effects of unilateral excision of the adrenal gland, section of the splanchnic nerves and section of the renal nerves on the secretion of the kidney. Their results were that all three operations increased the urine output as well as caused a more dilute urine to be secreted. They stated that the NERVE FIBERS IN THE RENAL CortTEX OF THE CAT 589 removal of the adrenal gland affected the kidney because the nerves going to the kidney in the vicinity of the excised gland had been injured. Kaufman and Gottlieb (1931) described a nerve supply to both the renal parenchyma and the vascular network. They also noted a rich supply of nerves in the fibrous capsule of the kidney. They observed nodular, brush-like and arborized nerve endings in the kidney tissue. Nerves were observed terminating on both tubules and arteries with some fibers being traced to the capillaries ending in the glomeruli. They believed a greater number of tubular nerve fibers existed but these fibers could not be demonstrated due to the methods of staining used. Gruber (1933) summarized from the accumulated literature and by some of his own experimentation that the kidneys are innervated by parasympathetic and sympathetic nervous elements. He proposed that the functions of the sympathetics are both vasodilata- tion and vasoconstriction, these actions being sufficient to explain the secretion of urine without the assumption of specific secretory nerve fibers existing in the kidney. We believed that the function of the parasympathetic fibers had not been proved satisfactorily either by physiological or pharmocological investigations. Also from a study of the literature, and some of his research, Szabo (1948) concluded that the arterial network (including the glomeruli) and the tubular epithelium, the two most important formations of the kidney, have nervous connections. The practical significance of these facts, he points out, is that sensory fibers terminating in the capsula fibrosa and along the arteries register local pressure conditions under normal circumstances. He postu- lated a reflex arc whose efferent fibers cause vasoconstriction in the renal arterial network. Vasoconstriction in turn cause a de- crease or possibly a loss in glomerular filtration. With regard to renal sympathectomy Szabo concluded that it is actually a periar- terial sympathectomy, that is, a denervation of the kidney’s arterial network. The denervation of the functional arterial network influ- ences glomerular filtration as discussed above. Using Bodian’s Protargol method, a more specific staining tech- nique for nerves, Harman and Davies (1949) studied the intrinsic innervation of the mammalian kidney. They found that nerves and nerve endings were more numerous than had been previously reported. They observed nerve endings on the adventitia of 590 THe UNIVERSITY SCIENCE BULLETIN blood vessels, the epithelium of the pelvis, between tubular cells and in the perivascular tissue of the glomeruli. The structure of most of the nerve terminals they noted were fusiform, club-shaped and ring-shaped. They proposed that the free nature of the nerve terminals and the character of the glomerular complex suggests an afferent mechanism thus implying that the kidney is vast sensory organ. Christensen, et. al., (1951) utilized degeneration studies to state that most of the nerves in the kidney are associated with arteries and arterioles. They supported the concept that the efferent innervation of the renal arteries, veins, and the rest of the vascular bed in the kidney is solely sympathetic, and that there are no parasympathetic nerves in the perivascular plexus within the kidney. Physiological experiments were also conducted in order to con- tribute information concerning the actions of the renal nerves in relation to functioning of the kidney. Kaplan, et al., (1951) showed that denervated kidney of a dog demonstrated a higher concentration of sodium and chloride in the urine flow. Block, et al., (1952) demonstrated the fact that electrical stimulation of the renal nerves under optimum conditions caused sufficient renal vasoconstriction to result in the cessation of the blood flow through the kidney for a few minutes. Hoff, et al., (1951) described the results of electrical stimulation of foci on the anterior sigmoid gyrus of the cerebral cortex in cats. As a result of stimulation they observed a transient elevation in blood pressure in the general circulatory system together with renal cortical ischemia. Unilateral renal denervation essentially abolished these ischemic changes in the denervated kidney during stimulation of the brain. Mitchell (1951) conducted extensive morphological studies of the intrinsic nerves of fetal and adult human, cat, rat, macaque monkey and mouse kidneys. With a variety of techniques he stained kidney sections for both nervous and connective tissue ele- ments. He established a table of criteria for distinguishing the nerves from the connective tissue strands. He pointed out that the existence of rich intrarenal nerve plexuses, of nerve plexuses within the glomeruli, of distinctly myelinated nerve fibers within the kidney and interarenal ganglia or nerve cells could not be confirmed. He demonstrated nerve filaments supplying both the tubules and the glomeruli. Mitchell did not distinguish sym- pathetic from parasympathetic nerves in the kidney. NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 59] Kuntz (1953) stated that afferent fibers of both spinal and vagal nerves traverse the renal plexus. Preganglionic sympathetic fibers traverse the splanchnic nerves passing either through the coeliac plexus or directly joining the renal plexus. He states that nerve fibers of vagal origin have been traced into the renal plexus by various investigators but these parasympathetic fibers probably do not extend into the renal parenchyma. Maillet (1959) worked with frogs, salamanders and guinea pigs describing nerve fibers going to the tubules, to the capsule of the glomeruli and entering each glomerulus to end on the cellular membrane of the floccular capillaries. He concluded that the sympathetic role in the kidney involves glomerular control, arterial control and partial tubular control. He suggested that the para- sympathetics have as their function tubular control. The investigations cited indicate the presence of sympathetic nerve fibers in the kidney but do not establish the presence of para- sympathetic nerves in the kidney. In the present investigation denervated kidneys of cats were utilized to derive more informa- tion concerning the presence of sympathetic and parasympathetic nerves in the renal cortex of cats. The absence of nerves in the denervated kidneys contrasted with the presence of nerves in the normal kidneys will establish the presence of sympathetic and/or parasympathetic nerves in the renal parenchyma. It is believed that there are parasympathetic nerves in the kidney but investiga- tions to date have not been able to give conclusive evidence of their presence. MATERIALS AND METHODS Nine apparently healthy and normal male and female adult cats of approximately the same size were used for these experiments. Three of the cats were sacrificed without being subjected to any operation. An operation was performed on three cats in order to remove the right half of the coeliac ganglion and plexus. The remaining three cats were subjected to removal of a cervical seg- ment of the right vagus nerve. The cats in which no operation was performed have their kidneys designated as normal kidneys. In the operated animals the right kidneys are designated as ex- perimental and the left kidneys are designated as control. After sacrificing the cats, the kidneys were removed from the body, the capsula fibrosa was striped off the kidney and the kidney was cut longitudinally into symmetrical halves and immersed into ten per 592 THe UNIVERSITY SCIENCE BULLETIN cent formalin. The anaesthesia used in the necessary operations was veterinary nebutal in dosages of one milliliter per five pounds of body weight. Samples of the kidney tissues were imbedded in Fisher tissuemat (56°-58°), sectioned at twelve to fifteen micra and mounted on glass slides. Three stains were employed to try to bring out the nervous elements; these were Holmes (1943), Brown-Vogelaar (1956) and Peters (1958). The performance of both the Holmes stain and the Peters stain was adequate for staining nerve bundles but was inadequate for staining the finer nerve fibers. The per- formance of the Brown-Vogelaar stain gave adequate results for the histological studies. The purpose of choosing the coeliac plexus for removal was to cut off the major sympathetic and parasympathetic nerve supply to the kidney. The right side of the animals was chosen at random as the experimental side. The removal of the vagus from the neck, due to its relative accessibility, was preferred to removal of the vagus infradiaphragmatically. In counting the nerve fibers in a given histological area in the kidney cortex, one should theoreti- cally find fewer fibers in the coeliac denervations than in the vagal denervations since it is cited in the literature that there are sym- pathetics but no parasympathetics in the renal cortex. To remove the right half of the coeliac ganglion in the given cats, the operation commenced by entering the abdominal wall at the linea alba. The intestines were displaced to the outside of the body and kept warm and moist with a 0.9 per cent saline solution. In the three cats involved in this operation it was noted that both halves of the coeliac ganglion were closely associated. The ganglion was usually found in the mesentery proper anterior and partly dorsal to the renal artery (near its branching from the abdominal aorta) and medial to the hepatic portal vein. One half of the ganglion leading to the right side and any branches of it going to the right renal plexus were removed. The intestines were replaced and the abdominal wall was closed. After allowing two to three weeks for degeneration the animals were sacrificed. Before the kidneys were removed the right side was checked to be certain that the connections of the coeliac ganglion had not regenerated. As far as could be determined by gross inspection the connections had not been reestablished. In the vagotomy operations an incision was made in the skin between the outlines of the sternocleidomastoid muscle and the sternohyoid muscle on the right side of the neck. After exposing NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 593 the carotid sheath in the underlying fascial planes, the vagus nerve was separated from the cervical sympathetic chain. A 1.5 to 2.0 centimeter section of the vagus nerve was excised. The structures were replaced in their normal positions and the skin was closed. Approximately four weeks were allowed for degeneration in these animals, After this time the animals were sacrificed and checked to be sure the section of the vagus had not regenerated. No re- generation was noted upon gross examination of the area. The time allowed for maximum degeneration in the experimental cats is given as follows: COELIAC REMOVED VAGUS REMOVED Cat No. Days Cat No. Days III 19 VI 31 IV 18 VII 30 V Ohl VIII 29 According to Cajal (1959) degeneration takes place in fibers of medium thickness in fifteen days and in the finest nerves in nine to ten days with degeneration occurring simultaneously along the entire length of the fiber. The amino-silver method of Brown and Vogelaar (1956) was used to stain the kidney sections. This staining method is a modi- fied Bodian technique used for 10 percent formalin fixed tissues imbedded in paraffin, sectioned and mounted on slides. The staining is accomplished by an amino acid-silver nitrate solution to which metallic copper strips are added. The amino acid (glycine) stabilizes the dilute silver nitrate solution. This im- pregnating solution is adjusted to a pH of 8.0-8.5 with normal sodium hydroxide. The slides to be stained are placed in this solution for 24-48 hours at 37 degrees centigrade. The remaining steps in the staining procedure are standard: the reduction, rins- ing, gold toning, oxalic acid treatment and sodium thiosulfate treatment follow the original Bodian technique. The silver impregnanted tissues were used to determine differ- ences in the quantitative distribution of nerve fibers in the renal cortex. The data of Table I deal with the presence of fibers around the renal tubules; those of Table II deal with the presence of fibers around the glomeruli; while the data of Table III deal with the distribution of nerve bundles in the tissue sections. Generally speaking the studies are designed to demonstrate either the morpho- logical differences or simularities between the normal, experimental and control kidneys. This information could contribute to the determination of the presence of parasympathetic and/or sym- pathetic nerves in the cortex of the cat kidney. 594 THe UNIVERSITY SCIENCE BULLETIN The fibers referred to in Table I were the black strands observed around the convoluted tubules in the renal columns of the cortex. (See photographs 7 and 8). Since the staining did not allow a distinction to be made between proximal and distal tubules no effort was made to include this information in the tables. The fibers surrounding both longitudinal and cross sections of tubules are included in the data. The fibers referred to in Table II were those observed surrounding the capsular layer of the renal corpuscle. (See photographs 5 and 6). Random fields were brought into focus and the tubules or glom- eruli and their fibers were counted and measured. All of the studies were done with an ordinary light microscope using a magnification of approximately 440 x. The values assigned to the length of the fibers around either the tubules or the glomeruli in Tables I and II are as follows: Strands\ completely, around =.) .-.5.4. . 4 ake 5 Strands: three-quarters around {332} 3.25205 ee 4 Strands: ‘one-half around: .9:0.. 24s oe eee 3 Strands one-quarter around —) 0 2.0.6). Gee 2) Strands one-eighth around’ 29. 220224. 22) ae ee 1 Strands .not--6vident .e.0ceh 4 se atin sn Es A 8 eae err 0 Each individual value was then tabulated and recorded as the total length of strands (T). The tubules or glomeruli whose strands had been measured, were tabulated and recorded as the total number of tubules (N) and the total number of glomeruli (G). The total length of the strands was divided by the total number of tubules or glomeruli to give the mean length of the strands per tubule (T/N) or per glomeruli (T/G). The per cent difference between the mean length of the strands for each pair of kidneys from the same cat was calculated and entered in the tables. The study of the distribution of nerve bundles observed in the tissue sections is given in Table III. The information concerning the location of the observed nerve bundles and the structures to which they are adjacent is of interest in this table. (See photographs and: 2): The slides used in this experiment were labeled so that the Roman numeral designates the cat number. Cats I, II and IX are the normal cats which had not been subjected to any operation. Cats Ill, IV and V were those whose right coeliac plexus had been removed. Cats VI, VII and VIII were those who right vagus nerve in the cervical region had been sectioned. The capital letters N, NERVE FrBeRS IN THE RENAL CoRTEX OF THE CAT 595 E and C stand for normal, experimental and control, respectively. The A and B adjacent to the E and C in Cats III to VIII designates the sample of tissue cut for that given set of slides. The A and B following the N in Cats I, II and IX designates individual samples of tissue taken from the right or left kidneys of these cats. The Arabic numerals designate what slide of a given sample of tissue was used for the study. There are five to six sections of kidney tissue on each slide with most of the sections not in series. RESULTS AND DISCUSSION Three cats were employed for each of the groups of treatment, with a total of nine cats being used in these experiments. Three cats were used for each treatment for the purpose of giving con- vincing results than would be obtained by using a lesser number of animals. The right kidney of each cat was compared with its left kidney. The data collected for all nine animals are presented in Tables I, IJ and III. Panel A in the three tables includes the data pertinent to the normal cats which had not been subjected to an operation. Panel B in all three tables includes data related to the cats which had their coeliac ganglion and plexus removed. Panel C in all three tables includes data related to the cats which had a cervical segment of their vagus excised. In Panel A of Table I there is a per cent difference in the mean length of strands observed per tubule between the right and left kidney of each of the normal cats. The individual values ranged from 3.50 per cent to 9.18 per cent. Since all of these values are less than ten per cent they may be considered as normal variations from one animal to another. Panel B of Table I is concerned with the per cent decrease of the mean length of fibers between the right and left kidney of the cats that had undergone the operation for removal of the right coeliac. The right kidneys showed a per cent decrease in the mean length of strands per tubules from the left kidney in all three cases. The per cent decreases ranged from 17.30 per cent to 25.46 per cent. The low value (17.30 per cent for Cat III) is nearly twice the per cent difference of the highest value for the normal cats (9.18 per cent for Cat IX). The highest value for the coeliac denervated cats (25.46 per cent for Cat V) is nearly three times the value of the highest normal cat (9.18 per cent for Cat IX). The lowest value for the normal cats (3.50 per cent for Cat I) compared to the highest value of the experimental cats (25.46 per cent for Cat V) shows the per cent decrease in 596 THe UNIVERSITY SCIENCE BULLETIN the experimental kidneys to be eight times as much as the per cent difference in the normal kidneys. In comparing the right experi- mental kidney to the left control kidney, the kidneys of Cats III, IV and V show at least a reduction of two times and at most a reduction of eight times the per cent difference between the right and left kidneys of normal Cats I, II and IX. Panel C of Table I is concerned with the per cent decrease of the right experimental kidney from the left control kidney in the cats that had undergone the operation for removal of a cervical segment of the right vagus nerve. The right kidney shows a per cent decrease in the mean length of strands per tubules from the left kidney in all three cats. These per cent decreases ranged from 22.80 per cent to 36.94 per cent. The lowest value (22.80 per cent for Cat VII) for the experimental cats is more than two times the per cent difference of the highest value for the normal cats (9.18 per cent for Cat IX). The highest value for the vagal denervated cats (36.94 per cent for Cat VI) is four times the highest value (9.18 per cent for Cat IX) for the normal cats. The highest value of the vagal denervated cats (36.94 per cent for Cat VI) is thirteen times the lowest value (3.50 per cent for Cat I) for the normal cats. In comparing the per cent decreases of the mean length of strands per tubule between the experimental and normal animals, the right kidneys of the experimental cats show a reduction of two to thirteen times the difference between right and _ left kidneys of normal cats in the mean length of strands per tubule. The mean length of strands per tubule was effectively decreased in the right experimental kidneys of all three cats that had their right coeliac plexus removed and in the right experimental kidneys of all three cats that had a cervical segment of their right vagus nerve removed. This decrease in the mean length of strands per tubule is not attributed to a loss of connective tissue elements around the tubules but to the loss of nerve fibers. If this loss represented a connective tissue loss, this would be of extreme in- terest due to the fact that there is no great loss of connective tissue resulting from the denervation of an organ. The data showed that removal of the right coeliac plexus and the right cervical vagus results in degeneration of some, if not most, of the nerve fibers around the renal cortical tubules. The conclusion derived from the data of Table I is that there are nerve fibers around the renal tubules that have originated from the coeliac plexus and the vagus nerve. NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 597 TABLE I.—A Comparative Study of the Fibers Observed Around the Tubules in the Cortex of Cat Kidney es Per cent SLIDE NUMBER T N T/N Ae eee PaNEL A: NoRMAL Cats 5 INGA Deis ese kt ymnier veer an eial une RY oy Natt 706 225 | 3.14 3.50 ESN = De Giae Fee Tne ee en one aicaNinnri gles Sat ys 383 123 | 3.03 TST INVA 2 OI a eee Oa a te ta bee 340 117 | 2.91 6.43 WTEINIB ED ee sauce esa ae en 246 79 | 3.11 EXECS NGAS OES eg es na hina be eee tgs 338 115 | 2.94 9.18 IDS SINGT BES ite eae repeats Scie p rian cotta dees epeanter 353 132 | 2.67 PaNEL B: CoEeuiac DENERVATED CaTs STS A Ole amck eek ems bie ee el ere ey 621 317 | 1.96 17.30 TPIS CAS Siar acme ero nigh aa eee ts ce Mnee ae 670 DSam lel mon TOYS NSS eet as Wee bo er pine Seieciee Cl, aa meee Wren 193 89 | 2.17 24.39 TENE Ay eee fol tere near gp ite a Reet ame 162 56 | 2.87 EVES HANS yee aegis ca eauitrmetr a ey Ne ital rae h attaati bc ries 287 102 | 2.81 25.46 WES CBSA May arse nine share ematical ate 9? 617 161 | 3.77 PANEL C: VAG@ECTOMIZED Cats NYG LS) DNR a gi cain ae avtaate cunt Na seine uh eiacnaniie a Mla 244 144 | 1.69 36.94 AV Be Cy SNe rls aati ata tar er een eae ee rene te aap meee 332 139 | 2.68 AWA OS Ba oi ewe ign eh neta aan arma sok ntuRtl 376 146 | 2.37 22.80 WATE CASS is. a ean eS ae nn 621 202 | 3.07 WaT EB ee eR ER ake Rise Ti 404 138 | 2.93 30.07 WITTE CARB eee eee ie ee ater an 699 167 | 4.19 In Panel A of Table II there is a per cent difference between the right and left kidneys of each of the normal cats in the mean length of strands observed per glomerulus. These per cent differences ranged from 6.97 per cent for Cat I to 10.24 per cent for Cat II. These values are ten per cent or less and one can expect such a fluctuation to represent normal variation among individuals. Panel B of Table II shows a per cent decrease in the mean length of strands per glomerulus of the right experimental kidney from the left control kidney in all three cats which had undergone an opera- tion for removal of the right coeliac ganglion and plexus. These values ranged from 17.55 per cent to 26.96 per cent. The lowest value (17.55 per cent for Cat V) in the experimental cats is almost two times the highest value (10.24 per cent for Cat II) for the normal cats. The highest value for the experiment cats (26.96 per cent for Cat III) is nearly three times the highest value (10.24 24— 1367 598 THE UNIversitry SCIENCE BULLETIN per cent for Cat II) for the normal cats. The highest value for the experimental cats (26.96 per cent for Cat III) is nearly four times the lowest value (6.97 per cent for Cat IX) for the normal cats. In Cats III, [V and V the per cent decreases from the left con- trol kidneys to the right experimental kidneys are two to three times the per cent differences between the right and left kidneys of the normal cats. The facts show there had been removal of nerve fibers around the glomeruli of the renal cortex in the right coeliac denervated cats since the mean length of strands observed per glomerulus was effectively decreased in the right kidneys of these coeliac denervated cats. This decease in the mean length of strands around the glomerulus is not attributed to a loss of con- nective tissue elements but is attributed to a loss of nervous tissue elements around the glomerulus. If this were a connective tissue loss this would be of extreme interest due to the fact that one does not usually find a loss of connective tissue as a result of denervation of an organ, In Panel C of Table II a per cent decrease in the mean length of strands around the glomerulus from the left control kidney to the right experimental kidney is evident for Cat VI of the right cervically vagotomized cats. This value is 27.14 per cent. The per cent decreases noted for Cats VII and VIII were ten per cent or less which is within the range of normal variation. Cat VI shows a per cent decrease (27.14 per cent) which is comparable to the highest value (26.96 per cent for Cat III) of the per cent decreases in the cats in Panel B of Table II. The data of Table II showed that removal of the right coeliac plexus results in degeneration of some of the nerve fibers around the glomeruli. Due to the contradictory results concerning the removal of the right cervical vagus, the presence or absence of nerve fibers from the right vagus around the glomeruli could not be determined. The conclusion derived from Table II is that there are nerve fibers around the renal glomeruli that have orignated from the coeliac plexus. No significant statement can be made regarding the absence or presence of fibers which originated from the right vagus nerve around the glomeruli at this time. Table III is concerned with the mode of entry of nerve bundles into the renal cortex. The total number of nerve bundles observed in the tissue sections of the normal, experimental and control kid- neys was sixty-five. Of these five were observed in the medullary NERVE FIBERS IN THE RENAL CorRTEX OF THE CAT 599 TaBLE II.—A Comparative Study of the Fibers Observed Around the Glomeruli in the Cortex of Cat Kidney : Per cent SLIDE NuMBER T G T/G deareace PanEL A: NoRMAL Cats ITESINGANSS Ieee Sata Oat ern eta Nea bera ye Ri 311 101 | 3.08 8.76 VEIN Bae ioe eae ge ae ane emo rere eR Nn 247 88 | 2.81 ATES NGA eect ale hoes eat a eng RMR eau Ue Reo ei 132 52 | 2.54 10.24 TE SN Br ie eee Oe rac wes en ea Iii eee 162 WM | BS EXE INGA Geel eae aise ey SES ANE aay ae 183 oleae 5S 6.97 XEN Se rare Se rice es oe ea ee 252 105 | 2.40 PaneEt B: Coeuiac DENERVATED CaTS 1) ees) oy Nn Esa NINE ocr acl gee ee te a 430 201 | 2.14 26.96 JES Co) Bal US 7 en oR UE Rae nn en 396 135 | 2.93 VEY BOs SG Sa is ca Oh aloo re eC 695 256 | 2.72 21.38 TIVE CAMO B= Oy aie lice wee a so meas renege tes 1173 399 | 3.46 WE WAS De HIB Oye aero toe amancve ty ool 643 217 | 2.96 17.55 Wis CARR © IS SANG Wnt ee at yt an een 876 244 | 3.59 PaNEL C: VAGECTOMIZED CaTs WASH ASIIB 3) 2 Oh Ree ee eR Uva eee es 679 HRS, || Poe 27.14 WAS OAC BeBe otis coitus: Co aby Sas Shc tans Hiebert 858 253 | 3.39 WVEEHWA-6, -HB-2). 0 i ene a eee 597 192 | 3.11 10.11 eV aT = @ Ate CB 08 Ae Se AON OG fu heat iiie 3 ee aah t 948 274 | 3.46 Wel BA HIRED ieee ke ences ace 806 233 | 3.46 8.71 WATE CAVE CB Be en tes eee coer Rat as. oi Ga 780 206 | 3.79 tissue and sixty were observed in the juxtamedullary tissue. Of all sixty-five of the nerve bundles observed, thirty-two were ad- jacent to blood vessels and thirty-three were adjacent to renal tubules. Most of the nerve bundles noted in the juxtamedullary region of the kidney were observed in the renal columns rather than in the medullary rays. The significance of the data collected in Table III is that nerve bundles enter the renal cortex via blood vessels and between tubules apparently with equal frequency in the kidneys of both the coeliac and vagal denervated cats. The coeliac ganglion in the cat, according to Reighard and Jennings (1961), contains postganglionic neurons which synapse with some of the preganglionic fibers from the greater and lesser splanchnic nerves. The postganglionic fibers from the coeliac ganglion are accompanied by fibers from the vagus nerve which also sends a supply of fibers to the coeliac plexus. Some of these vagal and coeliac plexus fibers then proceed to the renal plexus to 600 TaBLE III.—Distribution of Nerve Bundles in the Control and THe UNIVERSITY SCIENCE BULLETIN Experimental Kidneys Number Bundles SLIDE of ob- Location of adjacent to: NUMBER served bundles Blood | Renal ESIVES vessels | tubules PaNnEL A: NoRMAL Cats TSNIAS Drs eee oa ey pe 3 1 medullary 2 juxtamedullary 2 1 T=N Ba2 8s Gate wae ee 5 5 juxtamedullary 3 2 PANEL B: CoELiac DENERVATED CaTs SHAS Ay serie ee 1 1 medullary 1 (hilus) TTI-CB=32 3c eek le, 1 1 juxtamedullary 1 TV CARS pier ie Gendt iss 5 1 medullary 4 juxtamedullary 5 DV CBO a arate ors es 8 8 juxtamedullary 3 5 WVEAS ORs eaten eee nema eels 1 1 juxtamedullary 1 NERDS cs fc ne ae Sree 4 4 juxtamedullary 2 2 VEC AA ee iS orcas. asap aku 1 1 juxtamedullary 1 WSO Baas oeteies Grses et sal 4 4 juxtamedullary 3 1 PANEL C: VAGECTOMIZED CaTSs WALSH Ar oie eisorarean te byte: mee 6 6 juxtamedullary 2 4 WAH CASS 3) SE eee de poet 6 5 juxtamedullary 1 medullary 1 5 Vi C Bas acaaa h ie ae 4 1 medullary 3 juxtamedullary 2 2 VASE B= 2a ee yr eee 2 2 juxtamedullary 2 NA CAH 3. co cira ta ieee eer: 2 2 juxtamedullary 1 1 WA © Bote inne yet aoe 4 3 juxtamedullary 1 juxtamedullary 4 (hilus) VASA S23 eect anes 2 2 juxtamedullary 2 VIER Ba 2n ccc een: 1 1 juxtamedullary 1 VA CAH 3 tea ler icgen eee 3 3 juxtamedullary 2 1 Vili -CBa3 2 ae 2 2 juxtamedullary 1 1 TRODATS 2 Riese en ne on eas pea 65 5 medullary 32 33 60 juxta- medullary NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 601 supply the kidney. Kuntz (1953) points out that there may be vagal fibers reaching the renal plexus by some other route than via the coeliac plexus. In our experiments the results show that sympathetic and para- sympathetic fibers are present around renal cortical tubules as concluded from the data of Table I. Table II shows that there are sympathetic fibers present around the glomeruli but the presence or absence of parasympathetic fibers could not be definitely de- termined. Table IIT shows that nerve bundles enter the renal cortex via blood vessels and renal tubules with equal frequency. In the right coeliac denervated cats (III, IV and V) no attempt was made to distinguishd parasympathetic fibers from sympathetic fibers passing through the coeliac plexus. There might be some interchange of fibers between the right and left coeliac ganglia but a great interplay of fibers was not observed in the animals before removal of these plexuses. In the right vagotomized cats (VI, VII and VIII) no attempt was made to determine the contributions of the left vagus nerve to the kidneys. Experiments on the right vagus nerve showed that it contributed to the innervation of the kidneys. Considering those kidneys that were functionally con- trolled by parasympathetic innervation, one would naturally sup- pose that right vagal fibers would enter the right renal plexus and left vagal fibers would enter the left renal plexus. However, in development there might well be some interchange from one side to the other. The results of these experiments with regard to the presence of sympathetic and parasympathetic nerves in the cat kidneys seem to support Gruber’s (1933) concept that the kidneys are innervated by both sympathetic and parasympathetic nerves. The results seem to support Maillet’s (1959) hypothesis that the sympathetics are involved in glomerular and partial tubular control and that the parasympathetics are involved in tubular control. The results are in opposition to Christensen, et al. (1951) who maintained the concept that the kidney is solely sympathetic with no parasym- pathetic contributions being present in the renal parenchyma. The results concerning the equal frequency of nerve bundles observed adjacent to blood vessels and renal tubules also is in opposition with Christensen, et al., (1951) who proposed that nerves enter the renal cortex only along the arterial system. We wish to point out that the use of three cats for each experi- mental group, that is, normal, coeliac denervated and vagal 24a—1367 602 THe UNIVERSITY SCIENCE BULLETIN denervated, does not give enough information for statistical evalua- tion. The data collected in these experiments is not sufficient to make any final conclusions but does give convincing evidence of a definite trend regarding the presence of symathetic and para- sympathetic nerves in the renal cortex of the cat. SUMMARY AND CONCLUSIONS Silver impregnanted tissues were used for the purpose of de- termining the presence of sympathetic and parasympathetic nerve fibers in the cortex of the cat kidney. The tissues were obtained from normal and experimental animals. Both right and left kidneys of three cats which had not been operated upon were used as normal controls. Three cats were operated upon for the removal of the right coeliac ganglion and plexus and three other cats were operated upon for removal of a cervical segment of the right vagus nerve. The right kidney of each of these cats was designated as the experimental kidney and the left kidney of each of these cats was designated as the control kidney. Removal of the coeliac ganglion severed the major source of sympathetic nerves (greater and lesser splanchnic postganglionics ) to the renal plexus. Removal of the coeliac ganglion also included severing some parasympathetic fibers of the vagus nerve. Removal of the cervical segment of the vagus nerve severed a major source of parasympathetic fibers to the kidney. The data collected in the experiments presented in this study contributed the following information: 1. There are both sympathetic and parasympathetic fibers in the cortex of the cat kidney as determined by denervation studies of nine cats. 2. Sympathetic and parasympathetic fibers seem to be present in relation to the renal tubules of the cortex of the cat. 3. Sympathetic fibers seem to be present around the glomeruli of the cortex but no significant statement can be made about the presence of parasym- pathetics around the glomeruli. 4, Nerve bundles are observed adjacent to blood vessels and renal tubules in the cat kidney apparently with equal frequency. This demonstrates that nerves follow both the blood vessels and renal tubules into the cortex of the kidney. This study did not elaborate upon nor refute the heretofore cited facts that many sympathetic fibers traverse the coeliac and renal plexuses to arrive at the kidney, that parasympathetic fibers travel via the vagus nerve to the coeliac and renal plexuses to arrive at the kidney, that the normal contributions to the renal plexus are from the following structures: NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 603 chiefly from the coeliac plexus second and third lumbar segments of the sympathetic trunk aortico-renal ganglion aortic plexus moo to The results of the experiments with regard to the presence of sympathetic and parasympathetic nerves in the cat support the concept that both sympathetics and parasympathetics contribute to the innervation of the cat kidney. The results do not support the concept that there are sympathetic but no parasympathetic fibers in the kidney. The results also propose that nerves enter the renal cortex by following renal tubules as well as blood vessels. 604 THe UNIvERSITY SCIENCE BULLETIN LITERATURE CITED AZOULAY, L. 1894. C. R. Soc. de Biol., vol. 46, p. 336-338 cited in G, A. G. Mitchell (1951). BEALE, L. S. 1869. Kidney Diseases, Urinary Deposits and Calculous Disorders, Their Nature and Treatment. 8rd. ed., J. Churchill and Sons, London, cited in Mitchell (1951). BERKLEY, H. J. 1893. The intrinsic nerves of the kidney. A _ histological study. Bull. Johns Hopkins Hosp., vol. 4, p. 18. Brock, M. A., K. G. Waxm and F, C. MANnn 1952. Circulation through the kidney during stimulation of the renal nerves. Am. J. of Physiol., vol. 169:3, pp. 659-669. 1952. Renal function during stimulation of the renal nerves. Am. J. Physiol., vol. 169:3, pp. 670-677. BRADFORD, J. R. 1888. Innervation of the renal blood vessels. Proc. Roy. Soc., vol. 45, pp. 362-268. Brown, J. O., and J. P. M. VoGELAAR 1956. Amino-silver staining of nervous tissue. Stain Tech., vol. 31, pp. 159-165. Caja, S. Ramon Y. 1959. Degeneration and Regeneration of the Nervous System. Translated and edited by Raoul M. May. Vol. I, pp. 75-76, Hafner Pub. Co., N. Y. CHRISTENSEN, K., E. Lewis and A. Kuntz 1951. Innervation of renal blood vessels in the cat. J. Comp. Neur., vol. 95, pp. 373-378. GruBer, C. M. 1933. The automatic innervation of the genitorinary system. Physiol. Rev., vol. 18, p. 497. Harman, P. J., and H. Davies 1948. Intrinsic nerves in the mammalian kidney. Part I: Anatomy in mouse, rat, cat and macaque. J. Comp. Neur., vol. 84, pp. 225-244. Horr, E. C., J. F. Ket, Jn., N. Hastincs, D. M. SHoies and E. H. Gray 1951. Vasomotor, cellular and functional changes produced in the kidney by brain stimulation. J. of Neurophysiol., vol. 14, pp. 317-332. Howsrook, N. L. 1883. The termination of nerves in the kidney. Proc. Am. Soc. Micr., vol. 6, pp. 51-58. HouMEs, W. 1943. Silver staining of nerve axons in paraffin sections. Anat. Rec., vol. 86, pp. 157-187. Kaptan, S. A., S. J. Fomon and S. Rapoport 1951. Effect of splanchnic nerve division on urinary excretion of electro- lytes during mannitol loading in hydropenic dogs. Am. J. Physiol., vol. 166, pp. 641-648. NERVE FIBERS IN THE RENAL CORTEX OF THE CAT 605 KaAuFMAN, S, and R. Gorries 1931. Innervation of the renal parenchyma. Am J. of Physiol., vol. 36, pp. 40-44. Kuntz, A. 1953. The Autonomic Nervous System. 4th ed., Lea and _ Febiger, Philadelphia, pp. 34-36, 269-275. Mater, M. 1959. Innervation sympathetique du rein: son role trophique. Acta Neuroveg., vol. 49, pp. 302-343. MircHe.xi, G. A. G. 1951. The intrinsic renal nerves. Acat Anat., vol. 13, pp. 1-15. PETERS, A. 1959. Staining of nervous tissue by protein-silver mixtures. Stain Tech., vol. 33, pp. 47-53. Quinsy, W. C. 1916. The function of the kidney when deprived of its nerves. J. Exper. Med., vol. 23, p. 535. Reicuarp, J. and H. S. JENNINGS 1961. Anatomy of the Cat. 3rd ed., Holt, Rinehart and Winston, N. Y., pp. 412-415. SzaBo, E. 1948. Innervation of the kidney and its practical significance. Acta Urol., vol. 2, pp. 31-39. KEY TO ABBREVIATIONS JN, a er SEES ks car ey ag eee ace arteriole IB Wage tes eee tenn eae utes oy ets ont blood vessel | DI Bisaity peek aera Stan Uae arab A dichotomous branching of the nerve fiber CR erin ee es nce wire Rcte glomerulus IV aie ape ieee tneee tench en Muy Ceca ages medullary ray IN ety reer Ben i aaa Coriano) at: nerve fiber INE ieee ares once nak ree ote Aa nerve bundle IR Gere yee ae ete eee Gea er NCO, renal column CS Ree a ese aes ca aeaea rye: eta Sertied black strand 606 THE UNIVERSITY SCIENCE BULLETIN PLATE I Innervation of Renal Cortex in the Cat Fic. 1—Nerve bundle adjacent to a blood vessel in the juxtamedullary tissue. 100 x Fic. 2—Nerve bundle between tubules of the renal columns in the juxta- medullary region. 440 x Fic. 8—Unidentifiable black strands along an arteriole leading to a glome- rulus in the cortical tissue. 100 x Fic. 4—Black strands around blood vessels of the cortex which could not be identified as nerve fibers without doubt. 100 x NERVE FIBERS IN THE RENAL CortTEXx OF THE Cat 607 PLATE I 608 Tuer UNIVERSITY SCIENCE BULLETIN PEATE i Innervation of Renal Cortex in the Cat Fic. 5—Glomerulus in cortex with nerve partly surrounding it and branch- ing to adjacent tubules. 440 x Fic. 6—Nerve bundle branching to adjacent glomerulus. 440 x Fic, 7—Dichotomous branching of nerve fiber between renal tubules of cortex. 440 x Fic. 8—Black strands between cortical tubules which were measured and tabulated for Table I. 440 x THE UNIVERSITY OF KANSAS SCIENCE BULLETIN VoL. XLV JUNE 7, 1965 [No. 9 The Serpents of Thailand and Adjacent Waters BY Epwarp H. TaytLor ApstTract: This review of the serpents of Thailand is based largely on the herpetological collections made in the years 1957-1961 in Thailand and ad- joining countries by the author. This fauna is treated under ten families, 58 genera, 159 species and subspecies; certain unnamed varieties are discussed. Many of the emendations of generic names have been disregarded in ac- cordance with an interpretation of the recent laws of nomenclature and Aplopeltura, Gerarda, Erpeton, Calliophis, Aipysurus are used instead of Haplopeltura, Gerardia, Herpeton, Callophis and Aepyurus. The most recent listing of Thai snakes is that of Taylor and Elbel (1958) in which 129 names appeared; certain ones have been dropped from the list and the following names are added: Typhlops khoratensis Taylor Typhlops trangensis Taylor Typhlops diardi Schlegel Typhlops siamensis Giinther Python curtus Schlegel Sibynophis melanocephalus Gray Gonyosoma floweri (Werner ) Gonyosoma jansenii elegans (Wemer) Elaphe taeniura ridleyi (Butler ) Oligodon cinereus swinhonis (Giinther ) Oligodon cinereus multifasciatus (Jan) Oligodon cyclurus superfluens subsp nov. Oligodon cyclurus smithi (Werner ) Liopeltis baliodeirus cochranae Taylor Calamaria leucocephala Duméril, Bibron and Duméril Calamaria uniformis M. Smith Boiga drapiezii (Boie) Boiga dendrophila melanota (Boulenger ) Boiga nigriceps (Ginther ) (609) 25—1367 610 THe UNIVERSITY SCIENCE BULLETIN Boiga jaspidea Duméril, Bibron and Duméril Boiga cynodon var. Calliophis macclellandii macclellandii Reinhardt Calliophis maculiceps malcolmi Taylor Calliophis gracilis Gray Astrotia stokesii (Gray ) Trimeresurus erythrurus Cantor Trimeresurus sumatranus (Raffles ) Trimeresurus stejnegeri K. Schmidt No evidence is at hand for the retention of Enhydris chinensis (Gray) in the list. Trimeresurus puniceus (Boie) of the former list is the same as Trimeresurus kanburiensis M. Smith. Natrix modesta (Gunther) is dropped, and Boiga d. dendrophila of the former list is the same as Boiga d. melanota (Boulenger). Most of the species considered are redescribed from Thai specimens and 125 figures, mostly photos of entire specimens in black and white, illustrate more than one hundred of the forms. SERPENTS OF THAILAND 611 TABLE OF CONTENTS PAGE [NAARODWGIOW——SIMGSTMBOWIS CAs ob uecocldnos dusedonbuoucoscecueoees ob oene 617 INCKNOWLEDGEMBNTS iy cracnin fcr cae nee de als Soe eae tea ennui n 620 HATS PROVINCESs (Chamewaits)ins) co). ne ata ee eee 618 CuNERAL: CONSIDERATION: ON SERPENTS) 052.5 900s 2 nes). Ve 621 PAX ONOMIC! CONSIDERIATIONG sees y eee ee re a ee ee 632 (CIGASS ORV IP TU TAMae eerie areata ces DENT EIA) a SERRE NESE pane NN 632 Subclass Symaptosaurlama ys ca pee tr: cme ees eae a eee 632 OUEUID EFS QIAGEN IAT i ates sent | ie a NSE Pe tins aay CO ea Ue 633 SubordenmSeRPE NMES pes cere stg ck eens Se aang noe ys aia non 633 HamulysebyphlopidaesBoulengerse esate y ne ny eres 634 Cenusekyphlons@Oppeles sss see es ee ea 634 diuphilopsalineatus Oller ey ree ye ere te 636 Typhlops flowert Boulenger............................ 637 Typhlops albiceps Boulenger........................-5- 639 Typhlops braminus (Daudin)...........-.........20-5. 642 iy phlopsikhonatens7staluaiyl Oleie niet ens tse 643 Haphlopsttnangensis lay lons tight a) 2 tee ean 646 dajphlopsidiarduschleveles gas) os er 648 ajphlopsenucilenaschiegeleaemers tae ener ne eae 652 dipnlopsisvamvens? sa Gxt CT. ssemenie tt nee arene eer teal 654 BannilyeAmlirdae sis Simait nniee eosin Meet ete enciet taney Creer (ate 655 GenussGylindnophis: \Wiaclers ai are en eerie 655 CHMMOT DOS FURWS SOUS ooo oaicaeongacgonscdcapoasde 656 Cylindrophis rufus rufus Schlegel... .................-. 656 HamulyexenopeltidaciCopesn wea is ie eee 658 Genus Xenopeltis Reinwardt in Boie....................... 660 Xenopeltis unicolor Reinwardt in Bole.................-. 660 amily a ord Aer Giraiyeecce a ker pars nate Mana oe CMe Se ce an ea ea 663 Subfamily Pythoninae Boulenger............................ 664 Genus2ythonsDavdinie ee ye eae nd ap 664 iRathonsmolunts «(hininaeus) pease 4 ea ees naar 664 Python molurus bivittatus Schlegel. ..................... 665 Python reticulatus (Schneider).....................-... 666 IP EOC: CUURLUS ISOUBUAs o bcc os 0 gsgo cee adndedsdeosushoee 670 Pythoncuntus-bnongersmar Stuble eee ne eer ee 670 HamilyaDipsadidae:Gunthenayrc eee re ree te ee 673 Subfamuilyaleareatinac aay] OTe ent en ee ra ee 674 GenuseParcaseWaclere ee m a es on Meee ina 674 Pareas margaritophorus (Jan)........0 00000 cee 675 OT CasHmaculanviuse ly thir ea reise ee ee errr 678 Pareas hamptoni (Boulenger)......................0--. 680 ROreasicaninatiuse (Ole) par ener ete ee ee 681 (Parecasilacyis«(BOIe) ae ene Sa ees es seal 685 Roreasimalaccanus) (Reters)) eee es sea oe 686 Genus A plopeltura Duméril and Bibron.................... 687 Alplopelmanbods (BO1le) ne eae ea cee 687 612 THe UNIVERSITY SCIENCE BULLETIN Family Colubridae Boulenger. ))>. .-752.-5.2- 29200 eee Subfamily X~.. = 42... -e ee ee Xenodermus javanicus, Reinhardt. .724. eee Subfamily Sibynophinae Dunn: «2 922-42. ee eee Genus Szbynophis Pitzingers 6 ae. a). ee Sibynophis melanocephalus Gray... ........0000 00000 Sibynophis triangularis Taylor and Elbel................ Sibynophis collars (Gray). = 922s eee Subfamily Acrochordinae Jam... 32)... ee ae eee Genus Alcrochordus Hornstedt =. 324-42 ah ee ee Acrochordus javanacus Hornstedt... 3 30 -=se ee ee Acrochordus granulatus (Schneider)..................... Subfamily. Colubrinae Cope. i. .22 Ce ee ee Genus Gonyosoma: Wiaglern.. cj... ne ee ee Gonyosoma flowert (Werner)... ......0 0000000 cee eee Gonyosoma janseni, Bleeker... 32.2. oe ee Gonyosoma jansenii elegans (Werner)................-.- Gonyosoma oxycephalum (Boie in Boie).................. Genus! Hlaphe Fitzinger: 3... 5.00). ose eee Elaphe taentura’ (Cope)in,. 2-2 oe eee eee Elaphe taeniura ridleyi (Butler)....................0.-. Elaphe-taenvura taenvura (Cope). 2.25.25 ae ee Elaphe radiata (Schlegel), . 3. . Shenae ee ee Elaphe porphyracea, (Cantor)... 256d ee Elaphe porphyracea porphyracea (Cantor)............... Elaphe flavolineata (Schlegel)... 2403.2 oie nee Genus Ptyas Witziniger 2). 6.5 s-so0 gh a ee Ptyas korrose(Schilegel) nyc tce enn ne Piyas mucosus (hinnaeus))\. 2.4: 2. 4 ee ee Ptyas carmatus Gunithern..& sa asna2 see Lycodon laoensis Gtinther.. 2.2.0.0... 0000 e ec eee Lycodon capucinus Bole Hs noes fe ee eee Lycodon fasciatus (Anderson)........... 00000000 eevee ee Genus Dinodon Duméril and Bibron...................... Dinodon septentrionalis (Glinther) .<. 9.).4.. 255 eae ae Genus Oligodon Boie in Bole. a: 5-2 .2- ns ee eee Oligodon cinereus Gunther, 7.0.42 = tao oe ee Oligodon cinereus swinhonis (Gtnther).................. Oligodon cinereus multifasciatus (Jan)..............+-00- Oligodon inornatus (Boulenger).. 0.0.0.0. ce eee Oligodon dorsolateralis Wall... ..........0000 000 e eee Oligodon cycliurus: (Canton) a. oe eee Oligodon cyclurus smithi (Werner).............00020005- Oligodon cyelurus superfluens subsp. NOv.............24-- Oligodon purpurascens (Schlegel). ................20.05. Oligodon purpurascens purpurascens (Schlegel)........... Oligodon quadrilineatus (Jan)... 2.0.0.0. eee 715 716 716 SERPENTS OF THAILAND 613 PAGE Olhigodon:barront Ve Sith han ee oe 776 Oligodon taeniatus Glinther.........0..... 000.000.0005. 779 Oligodonmoynsoni Wits Smaitlinee ta re ee en eee 781 GenusyiropeltissMitzimgenm tte ea an tee 783 Iiopeltis baliodeira (Boie in Boie)...................-.- 783 Liopeltis baliodeira cochranae Taylor.................++- 784 Evopeltisnscriptius: Ckneobald) iver wees 22 erly tee 786 Genus Dryocalamus Gtinther........0.....0..0.0.....04-.. 788 Dryocalamus davisoni (Blanford)...............5...54- 789 Genus Plagiopholis Boulenger...................-0.000055 792 Plagiopholis nuchalis (Boulenger)................--.-.- 792 GenusRseudorabdion Jam) es a ee 795 Pseudorabdion longiceps (Cantor)................22.--- 795 GenusiCalamarnia dole ots es ee eee es 797 Calamaria vermiformis Duméril, Bibron and Duméril.... . 797 Calamaria leucocephala Duméril, Bibron and Duméril..... 799 Calamaria uniformis M. Smith....................0.-. 802 Calamaria pavimentata Duméril, Bibron and Duméril.... . 804 Calamaria pavimentata pavimentata Duméril, Bibron and VE AON VOLS gi a cave eh oe ee a et aL ta ah NS em IGE 804 Calamaria siamensis Gtinther..................0-0-0005- 806 GenustAthaetulla Mimi sees Sine enya snare 808 Ahaetulla caudolineata Gray... 00... ee 809 Ahaetulla subocularis (Boulenger)....................--. 811 Aihactulla:jfonmosa: (BOs te ea) te ae 814 Ahaetulla cyanochloris (Wall)... 00.00.0000 0 000 e ee eee 815 Ahaetulla ahaetulla (Linnaeus).......................-. 816 Ahaetulla ahaetulla ahaetulla (Linnaeus)................. 816 SubfamulywNeatricinae:Stemegenscane 6 yn ea ee 820 Gems aN trea cers ae NE Ge a MERE ME I oe 820 Natrix percarinata (Boulenger)..................-.-5.- 821 Natrirktrvangulrgena. (EO) i = ene ye ea 822 Niatrrasdeschawumseev- May lOve a ois crte e ee o 825 Natrix groundwateri M. Smith......................... 827 Natmixunasn Giaidlaw)ors nse. ee a eee 829 Natrix piscator (Schneider).................0 0.0020 005. 831 Naima avipunctatas(elallowelll) ey ee 835 Genus Opisthotropis Ginther..................0.0022 2005. 837 Opisthotropis spencert M. Smith....................0... 839 Genus Rarahelicops Bourret seein sake Sine sae eae 840 Parahelicops boonsongi Taylor and Elbel................ 840 Genus Pseudoxenodon Boulenger....................00005: 841 ESCUdOLeNOdonNacropsa (livin) ager re aera eee te 843 Pseudoxenodon macrops macrops Blyth.................. 843 Genus Macrophisthodon Boulenger.....................005. 846 Macrophisthodon rhodomelas Boie...................5-. 846 Macrophisthodon flaviceps Duméril, Bibron and Duméril... 849 Genustiiiabdophiseuitzine er ean see ree eae ie rarer 850 Rhabdophis subminiatus (Schlegel)................0.00. 851 Rhabdophis subminiatus subminiatus (Schlegel).......... 851 614 THe UNIVERSITY SCIENCE BULLETIN PAGE Rhabdophis subminiatus helleri (KW. Schmidt) ............ 854 Rhabdophis nigrocinciius (Blyth) sa «sane eee 855 Rhabdophis stolatus (Linnaeus) 202). 444 ee eee eee 858 Rhabdophis'chrysargus’ (Boie)= 2.2 se eae 862 Subfamily. Boiginae ss. 2.6 Gade, (hee en eee eee 865 Genus Borga Bitzinieere 5 eee eee nee 865 Boigaidnapiezit (Bole) hacia n- a ae ae eee eee 867 Boiga multomaculata (Bole)... 4 eee oe 869 Boiga cyanea (Duméril, Bibron and Duméril)............ 872 Boia dendnophila (Bole) see. ean aoe ee ene 874 Boiga dendrophila melanota (Boulenger)................- 875 Boiga nioniceps-(Gunther) pan so one eer oo 877 Boiga jaspidea (Duméril, Bibron, and Duméril).......... 878 Boigatcynodon (Bole). S25 eee ee 881 Boiga: Cynodon Varies. <3) nos eee ee 883 Genus -Dryophis, Dalman-5 3. seo ee eee 884 Dryophis prasinus Boies, as ee 885 Dryophis mycterizans (Linnaeus)... ....0....00 00 cee eee 888 Dryophis nasitus (Gacepede)s aaa ce a ee eee 888 Genus Dryophiops Boulenger. 5. #2. 2-300 ae eee 891 Dryophiops rabescens) (Gray). .24 0 2 oe eee 892 Genus Psammodynastes Ginther..........0.. 000000000000 894 Psammodynastes pulverulentus (Boie).............+-0+5- 895 Genus Psammophis Mitzimger,. 2.) 42). act a eee 898 Psammophis condenarus (Merrem).....5....).55.5.5.55 898 Psammophis condenarus indochinensis M. Smith......... 898 Genus. Chrysopelea: Bole. ss. Sto act ae ee 900 Chrysopelea paradisy Bolen 23.23 4255-2 ee ee 901 Chrysopelea ornata: (Shaw). 2-2) eee 902 Subfamily Homalopsinae Boulenger.....................00-4- 905 Genus Hnhydris Sonnini and Latreille..................... 905 Emhydris plimbea (Bole) (8 isms sea: eee 906 Emhydris bocounti (Jai) et eae eee 909 Enhydris enhydris (Schneider). .............00000+00ee- 911 Enhydris smitht (Boulenger).............. 000000000 eee 914 Enhydris jagori (Peters)is. 2) sae See 917 Genus Homalopsis Kuhl and Van Hasselt.................. 919 Homalopsis buccata (Linnaeus)..............000000 eee 919 Genus Cerbenws Cuvier. ss 50 eee ee 923 Cerberus rhynchops (Schneider)................++++0--- 923 Genus: Bitia Gray cop ee eee 925 Bitvashydrovdes Graver 5 ae ee eee 926 Genus Gerarnda Grain. 5.ds Gis ee ee 927 Gerarda prevostiana (Kydoux and Gervais).............. 927 Genus H7petonsGlinthene <4 sacs ee eee 930 Erpeton tentaculatwm Waacépedess.. =.) ae eee 930 Grp etOm Spi eosen a cite tae ROR NS ee se 933 Bamily Blapidae-Boie=. 2.02.2 3: cies ee ee eee 935 Genus Bungarus Daudine a5) ee ee oe 936 Bungarusijlaviceps Reinhardt; eee ee eee 937 SERPENTS OF THAILAND 615 PAGE Bungarus flaviceps flaviceps Reinhardt.................. 937 Bungarus candidus (Linnaeus)............ 02.0.0 .00005 940 Bungarus fasciatus (Schneider)................-.020-5- 943 GenuseVajaswaurentiives ery ee omer ee eee 946 IN GAO TOGHIG) QUIVGVOEYEUIS)) 5.5/3 Gio gal euloa es eo oe cooks oe 20% 947 Naja naja kaouthia Lesson in Férussac................. 947 Genus Vendraspism zinger. heey a ais ee ele serene 956 Dendraspis hannah (Cantor).............0.020 000-005. 957 GenuseMiatvcona: Grays een cee ae ee ee eal cee 959 Maticora bivirgata (Boie).................2+-20 eee eee 960 Maticora bivirgata flaviceps (Cantor).................... 960 Maticora intestinalis (Laurenti)...............-...-00-. 962 Maticora intestinalis intestinalis (Laurenti)............-. 963 Genus! Callioniis Grayiagy a ee ee oe 965 Calliophis macclellandii (Reinhardt)..................-. 966 Calliophis macclellandii macclellandii Reinhardt.......... 966 Calliophissqnacilis Graver a ae 969 Calliophis hught Cochran................000020 200000 972 Calliophis maculiceps Gunther. 6 2 9250 se ne 972 Calliophis maculiceps maculiceps (Giinther)............. 973 Calliophis maculiceps malcolmi Taylor................-- 978 KNamily Hy drophidae Boles... 9515 eee ee es 980 Subfamulyi@aticaudinaew ee ae ken eed eee ae eee 980 Genussatzcanda Iaurentive a: vie secre ens 980 Laticauda colubrina (Schneider)................-...+--- 982 ancaudadlacaudatamnien ate oe eee erie ce 984 Genus Avpysurus bacepede. 45. ee fe) ne we ee 986 Aipysurus eydouxt (Gray)..... 2.0... e eee ee 987 Subfamuilyothydrophiimae.s eee eS a eee ee ae 989 Genus MWeriliaxGray se ia en ee ate eden en 991 KeriliaijerdonieGravinss ser heirs fetes ele 991 Kerilia jerdonii siamensis M. Smith.................... 992 GenustAlstrotia Pischen i ae eae ee ee oe oie 994 VAlstrotiazstokestis(Gray) eae ee ae 996 Genuspeolpophise Nees nit ne ee el erie ae 998 Kolpophis annandalet (Laidlaw)....................-.- 999 Genusperaescutata Walle ssc rere ee ee eee 1002 eR Qescuialawrpentia, (Schmidt) seis et ee ae eee te ae 1002 Genus nhyanina: Grays cans eee eles leas oe 1006 Emhiydninarschistosa) audintad. ae eno See earls ae 1006 Genus Acalyptophis Boulenger.................--+-+++.+-- 1010 Aicalyplophisiperonis (Mumeril) es ee ee a 1010 Genustliialassophis. no wee eres Weta veel ee 1011 Thalassophis anomalus (Schmidt).................++-- 1012 Genus Microcephalophis Lesson.................0.2+05-55 1014 MicrocephalophisignaciasiSnawelc 1 oe ee ee 1014 Microcephalophis gracilis gracilis (Shaw)............... 1014 GconuspRelanis yD audine ieee tac: cr setae eee er is 1019 Pelamis platurus (Linnaeus).....................0000- 1019 616 THe UNIverstry SCIENCE BULLETIN PAGE Genus Lap emis! Grays oct sis a cciaps abe okey elle ene eee 1023 BapemisshandiwickiiGraye naan settler eee 1023 Genus Hidrophis Watreiller s. > jc. csc eee lenenene nate 1027 Hydrophis cyanocinctus Daudin. ... 0.002.000 ee 1028 Hudnophis: ornatus (Gray). ia see er ee 1033 Hadrophisornatus onnatus Gray .9))1-)) eee eee 1033 Hydrophis caerulescens: (Shaw)... +24.) skis aoe 1034. Hydnophis' torquatus (Gunther, .).).0)7 see ee ene 1037 Hydrophis torquatus diadema Ginther................. 1038 Hydrophis torquatus aagaardi M. Smith................ 1039 Hydrophis klosst «3,3 eiisins eter rol eerie ee ee oe 1042 Hydrophis fascratus schneider... -..). eka ene 1047 Hydrophis fasciatus atriceps M. Smith........7.....25. 1048 Hydrophis brooke: Gunther... 2443)... eae Eee 1051 Hyrophis mamillaris (Daudin). 0.0.0.0... cc ee 1052 Family Viperidae Bonaparte... 2.2... ce tee cla es ee 1055 Genus Vipera Maurenti 5 6 26. aa ec cacao en eee 1055 Vipera russellz (Shaw) icc a. hee eee eee 1055 Vipera russella siamensis M. Smith.................... 1055 Family ‘Crotalidaesss4.355 dey escepeys is alaliei evens tente valet tae oe ee a 1057 Genus Agkistrodon (Palisot de Beauvois).................. 1057 Agkistrodon rhodostoma (I. Bole).................+... 1058 Genus: Drnimeresurnus Lacepedes. 225 es ae ae ee eee reticulatus Rostral and two upper labials on each side with pits.................. . Ventrals 245 to 275; subeaudals 58 to 75; large elongate serpent reaching a length of five to six meters...................-40-. molurus bivitattus Ventrals 160-175; subecaudals 28-32; a much smailer shorter thicker ser- pent reaching to a length of about 2.5 meters........ curtus brongersmat to iw) Python molurus ( Linnaeus ) Coluber molurus Linnaeus, Systema naturae, vol. 1, 1766, p. 387; ibid., 10th Ed., 1758, p. 225 (type-locality, India); Andersson, Bih, Sv. Vet. Akad. Stockholm, vol. 24, 4, 6, 1899, p. 35. Python molurus Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 246; Catalogue of the snakes in the British Museum (Natural History) 2nd Ed., vol. 1, 1893, pp. 87-88; Bourret, Les Ser- pents de l’Indochine, 1936, p. 18. SERPENTS OF THAILAND 665 Two subspecific forms are recognized, the typical one molurus molurus in Ceylon and peninsular India, and molurus bivitattus Schlegel found in the Indo-Chinese Subregion, including southern China and Thailand. Python molurus bivitattus Schlegel (Fig. 8a) Python bivitattus Schlegel (part.), Essai sur la physionomie des serpens, vol. 3, 1837, p. 408, pl. 15, figs. 1-4 (type-locality, restricted to Java). Python molurus bivittatus: Mertens, Abh. Senckenb. Nat. Ges., vol. 42, 1930, p. 287, pl. VIII (type-locality fixed as “Java”); Pope, The reptiles of China, Fic. 8a. Python m. molurus Linnaeus. (From Taylor, Univ. Kansas Sci. Bull., vol. 33, pt. 2, 1950, pl. 13, fig. 1.) 666 Tue UNIVERSITY SCIENCE BULLETIN 1935, pp. 72-75. pl. 5; Bourret, Les serpents de Indochine, 1936, pp. 19-20, fig. 9; M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 108-109. Diagnosis: Large serpents; labials separated from eye by sub- oculars; arrow-shaped mark on head distinct; girth proportionally greater than in reticulatus. Description of subspecies (composite ): Nostril at upper posterior part of a large anterior nasal; rostral broader than high with two deep diagonally placed pits; a pair of internasals forming a median suture each longer than wide; a pair of enlarged prefrontals; these followed by a second pair of prefrontals; frontal divided longitudi- nally, each part about as wide as a supraocular; parietals frag- mented; 12 supralabials first two bearing pits, separated from orbit by a row of suboculars; two preoculars; four loreals; two post- oculars; about 20 infralabials, only two touching first pair of chin- shields; four pairs of chinshields all very small, last pair widely separated from ventrals; pupil vertical; mental small, length of its labial border less than half that of rostral; scalerows around body (molurus) 61-75; ventrals, 242-265; subcaudals, 60-72. Color: Ground color of yellowish brown or grayish brown; a dorsal series of large quadrangular, often somewhat elongated, reddish-brown black-edged spots; on each side a series of smaller spots with light centers; large spear-shaped dark mark from nape to tip of snout with an elongate cream or fawn spot on back of occiput, bordered laterally by a stripe of ground color; a dark stripe from tip of snout through eye widening on temporal region; a small brownish bar below eye; venter and chin yellow-cream; sides brown-spotted. Measurements: Reaches a length of about 6.5 meters (fide M. Smith). Distribution: Malcolm Smith reports three specimens taken in Thailand from the provinces of Raheng, Lop Buri, and Chon Buri (Siracha). Elsewhere the species occurs in southern China, Hong Kong, Hainan, and the whole Indo-Chinese subregion. Python reticulatus (Schneider ) (Fig. 9) Boa reticulata Schneider, Historiae amphibiorum naturalis et literariae, fas- ciculus secundus, 1801, p. 264 (based on Seba, Thesaurus, vol. 1, pl. 62, fig. 2, and vol. 2, pl. 79, fig. 1, the type-locality not recorded ). Python reticulatus: Gray, Zoological miscellany, 1842, p. 44; Cantor, Joum. Asiat. Soc. Bengal, vol. 16, 1847, p. 902; Giinther, Reptiles of British India, 1864, p. 330; Boulenger, The fauna of British India . . . Reptilia and SERPENTS OF THAILAND 667 Batrachia, 1890, p. 246; Catalogue of the snakes in the British Museum, vol. 1, 1893, pp. 85-86; Hagenbeck, Spolia Zeylanica, 1905, p. 197; Journ. Bombay Nat. Hist. Soc., vol. 16, Aug. 1905, pp. 505-506; (data from pre- vious article); Sclater, List of the snakes in the Indian Museum, 1891, p. 4: Flower, Proc. Zool. Soc. London, 1899, p. 654; Ridley, Journ. Straits Branch Roy. Asiat. Soc., 1899, p. 196; Mocquard, ‘Les reptiles de VIndo- Chine, 1907, p. 42: Boulenger, A vertebrate fauna of the Malay Peninsula Ros Reptilia and Batrachia, 1912, pp. 107-108, fig. 32 (lower fig.); M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1 ne orm i 1914, pp. 9-10 (Bangkok): Gairdner, ibid., vol. 1, no. 2, Aug. igi "pp. 195- 126: M. Smith, ibid., vol. 1, no. 3, Mar. "1915, pp. 155, 187, 213; ibid., vol. 1, no. 4, Dec. 1915, p. 244 (Koh Chang); Werner, Arch. Nat. "Berlin, vol. 87, 1921, p. 236; Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 253; ibid., vol. 31, 1926, p. 84; Kopstein, Trop. Natuur., 1927, p. 65; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 11, 1937, p. 61; Bourret, Les serpents de l’Indo-Chine, 1936, p. 16, fig.; M. Smith, The fauna of British India including Ceylon and Burma. 92) Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 109-110. Boa rhombeata Schneider, Historiae amphibiorum naturalis et literariae . vol. 2, 1801, p. 266 (based on Seba, Thesaurus, vol. 2, pl. 80, fig. 1). Boa phrygia Shaw, General zoology, vol. 3, 1802, p. 348, pl. 97 (based on Seba, Thesaurus, vol. 1, pl. 62, fig. 2). Se javanica Shaw; General zoology, vol. 3, 1802, p. 441 (type-locality, ava). Diagnosis: Large serpents (28-30 feet); posterior head-scales small; three to five loreals; rostral with two pits, first four supra- labials with single pits; 70 to 80 scales around middle of body; no enlarged chinshields; two or three anterior-and five or six posterior infralabials with pits; subcaudals divided. Description of species (from No. 1692, Kanchanaburi): Rostral about as wide as high, narrowly visible above, with two large di- agonal pits reached by sutures from first labials; a pair of large nasals forming a median suture; nostril dorsal, in posterior part of scale, entered by a suture arising at first loreal; a pair of internasals more or less fused to nasals with a median suture; two pairs of prefrontals, second pair broken into several scales; two or three loreals; two preoculars, upper largest; frontal partially broken ante- riorly; supraoculars enlarged, irregular, broken on inner side; three postoculars; temporal scales small; supralabials, 13-14, first four with diagonal pits, seventh bordering eye; infralabials, 23-23; second to fourth or fifth with small pits. A mental groove but no well- defined chinshields. Pupil vertical. The parietal region covered with small scales; scales about neck at first ventral, 65; at mid-body, 72: before vent, 44. Ventrals, 322; subcaudals, 92 + 1 or 2; anal single. Color in life: Above, head uniform brown growing light gray or dirty white on supralabials. A narrow black line begins on snout and runs back to nuchal region where it widens and termi- 668 THE UNIVERSITY SCIENCE BULLETIN Fic. 9. Python reticulatus (Schneider). No. 34530. Pattani, Pattani province. Actual total length, 727 mm.; tail, 112 mm. nates three centimeters back of head; a narrow black line from eye to back of head; chin and throat white with some indefinite blackish markings on infralabials anteriorly. On each side of body a zigzag black stripe connecting across back with a similiar series on other side, enclosing along back a SERPENTS OF THAILAND 669 series of somewhat oval brown spots edged with light brown; below the zigzag line, a series of ivory to yellowish spots with black spots above and below them; a second series of ocellated spots directly below first series. More posteriorly the pattern takes on the appearance of a series of black rhombs, touching each other laterally, with a light brown edge about a darker center; ventrals throughout cream to ivory; enlarged scalerows near ventrals with black flecks and spots; sub- caudal area black and white spotted. Measurements in mm.: Total length, approximately 3560; snout to vent, 3084; tail, 476. Variation: The sixth or seventh supraocular may border eye. There are usually two anterior infralabials bearing pits and five or six posterior infralabials with larger pits. Usually the internasals are free from the nasals. The scalerows at the middle of the body vary from 70 to 80. The range in ventrals and subcaudals are respectively: 297 to 332; 75 to 102; subcaudals usually divided. Distribution: In Thailand the species is found throughout the country except perhaps in the northern parts. Elsewhere it occurs in Burma, much of Indo-China, Malaya, and the Indo-Australian Archipelago as far as the Philippines. Remarks: In literature this creature is widely known as the python and is probably the largest (longest) serpent known. Speci- mens measuring 25 feet are not especially rare and specimens reach- ing a length of 30 feet have been reported, perhaps authentically. The South American anaconda, also belonging to this family, closely approaches the size of the python, and where specimens of equal length are compared the anaconda may be the heavier and more bulky species. Most if not all of the Oriental members of the family Boiidae are oviparous, while the new world species are chiefly ovoviviparous. The python has been known to lay as many as a hundred eggs in a single clutch. The eggs are cared for, the female coiling about the eggs and remaining with them until the young are ready to hatch. The incubation period is said to range between 60 and 80 days. Mr. John Hagenbeck writes (loc. cit., 1905): “A gigantic python which arrived here last year from British North Borneo via Singa- pore laid about 100 eggs on the 28th of October, almost filling the 670 THE UNIVERSITY SCIENCE BULLETIN box in which she was kept. On the following morning she had col- lected the eggs by skillful movements of her body into a heap which she completely covered in such a manner that the weight of the body exerted no pressure upon the soft-shelled eggs. “During the period of incubation I offered ducks, fowls, and geese to the python but she refused all nourishment. “On the 14th of January she left the mass of eggs quite exposed, and I had given up hope of a successful issue when I discovered the first nestling with half its body emerging from the egg, into which, however, it retired again towards evening. “The period of incubation lasted exactly two-and-a-half months. The young measures from two, to two-and-a-half feet long. “Altogether forty-five young have been recovered from the clutch. “So far as can be ascertained, the parent python measures about 28 ft. in length and weighs 250 lbs.” Python curtus Schlegel Python curtus Schlegel, Dierentuin Gen. Nat. Amsterdam, Rept. 1872, p. 54; g. — (type-locality, Sumatra); Hubrecht, Notes Leyden Mus., vol. 1, 1879, p. 244 (between Padang and Indrapura, Sumatra); Boulenger, Cata- logue of the snakes in the British Museum (Natural History), vol. 1, 1893, pp. 89-90; Zeneck, Tiibinger Zool. Arbeit., vol. 3, no. 1, 1898, pp. 46-49, 140-143. Python breitensteini Steindachner, Sitzb. Akad. Wien, vol. 82, 1881, p. 267; Fischer, Arch. fiir Naturges., 1885, p. 68, pl. 5, fig. 5 (type-locality, Borneo). Aspidoboa curta: Sauvage, Bull. Soc. Philom, ser. 7, vol. 8, 1884, p. 148. Three subspecies have been recognized by Olive G. Stull, loc. cit., one of which occurs in Malaya and Thailand; this has been named as follows: Python curtus brongersmai Stull (Figs. 10, 11) Python curtus Blanford, Proc. Zool. Soc. London, 1881, p. 222 (Singapore); Boulenger, Proc. Zool. Soc. London, Nov. 5, 1889, pp. 432-483, pl. 45 (in color; specimen from Malacca, Malaya). Python curtus brongersmai Stull, Occ. Papers Boston Soc. Nat. Hist., vol. 8, Jan. 28, 1938, pp. 297-298 (type-locality, Singapore. ) Diagnosis: This form may be distinguished from Python curtus curtus Schlegel and from the Bornean form Python curtus breiten- steini Steindachner by the presence of two supraocular plates in- stead of one; by the entrance into the orbit of two supralabials as compared to none (usually); and a higher average number of ventrals. SERPENTS OF THAILAND 671 Description of species (composite): Rostral wider than high with two relatively deep transverse pits, the scale narrowly visible above; a pair of internasals, slightly longer than wide; two pairs of pre- frontals, anterior smaller but equally as wide, forming a suture with its fellow equally as long as posterior pair; frontal segmented longi- tudinally the two together smaller than a supraocular; three pairs of parietals, posterior shortest; two preoculars upper very large; a series of suboculars; ten loreals on each side; ten supralabials, fifth and sixth bordering eye, first and second with deep pits; infralabials 19, second to fourth and tenth to fourteenth with shallow pits (third to fifth and twelfth to fourteenth on left side). Fic. 10. Python curtus Schlegel. B.M. No. 86.5.15.35, Telang, S. E. Bomeo. A little less than natural size. (This figured specimen is from Bomeo and may or may not be the same subspecies as occurs in Malaya and Thailand) 672 THE UNIVERSITY SCIENCE BULLETIN Scalerow-formula: 51-55-33; ventrals, 174; caudals, 32 (divided). Color: Above on upper part of dorsum somewhat reddish or orange-red; head somewhat lavender above with light lines on labial edges and another line from behind eye joining labial lines at jaw-angle; a row of cream or fawn-colored spots or spots joined tending to form a median extremely irregular lighter line; a fawn line from head dorsolaterally soon broken into irregular spots that may be more or less connected; a median series of fawn spots on latter fifth of body with darker lateral spots that may be connected; ventral areas cream. Dentition: Mandibular teeth, 19; premaxillary, 2; maxillary, 19; palatine, 6; pterygoid, 13. Measurements in mm.: Total length of type, 1180; tail, 112. Variation: Stull (1938) states that: “Two specimens in the British Museum from Malacca and Kuala Lumpur, Malaya have the following scale counts (made by Dr. H. W. Parker): Scalerows at mid-body, 58, 54; ventrals, 170, 168; caudals, 28, 27; infralabials 19/19 (19-18); two preoculars; two postoculars, two supraoculars, and two supralabials entering the eye on each side.” Distribution: This form has been taken in Thailand only from the extreme southern part in Pattani. Elsewhere it is known from Singapore, the type-locality “Malacca,” and Selangor (Kuala Lum- pur). Remarks: The young specimen figured (fig. 10) is B. M. 86.5.15.35, Telang, S. E. Borneo. It probably represents the Bornean breiten- steini. There is but little color difference in the three forms. The distinctions between the three subspecies recognized by Stull have to do primarily with differences in scale-counts. Thus the ventrals of brongersmai average, 170.5, breitensteini, 162.5, curtus, 159.6, a total variation in averages of 11 ventral scales! There are other dif- ferences. The form brongersmai has the supraocular broken, and two supralabials bordering the orbit. The single Thai specimen of the peninsular form that I have seen (fig. 11) is a living one taken on Bukit Besar, Pattani and now in the collection of Nai Prayoon Kananuracks. It was not feasible to make scale-counts on the living specimen, and in consequence | am not certain that it agrees wholly with brongersmai. The adults of this species are especially thick-bodied in proportion to their length. The head is proportionally small and short. | SERPENTS OF THAILAND 673 Fic. 11. Python curtus brongersmai Stuhl. From a living specimen photo- graphed against the side of a box; Bukit Besar, Na Pradoo, Pattani, Pattani province. Length approximately 2.25 m. (Property of Nai Prayoon Kana- guracks, of Na Pradoo, Thailand. The lateral black marks on the side of the head are indistinguishable from the background making the head appear too narrow; see fig. 10.) Tweedie (1954) states that “The Short Python is not uncommon in Malaya and is even more attached to water than the Reticulated [python ].” Famity Diesapmar Gunther The generic name Amblycephalus Kuhl, (1822) applied to an oriental genus of snakes is untenable since it is preoccupied by Amblycephalus Zeder (1803), and cannot be used either as a generic name or used in forming a family name Amblycephalidae. The next generic name available is Pareas Wagler, and the resulting sub- family name would be Pareatinae. Dr. James Peters who has recent!y reviewed a group of Central and South American species of the subfamily Dipsadinae does not commit himself to designating the family to which this subfamily belongs; one may presume that it is considered as a member of the Colubridae or Dipsadidae. Concerning the Oriental group including the genera Pareas (1830) and Aplopeltura (1853) Malcolm Smith (1943) states: “Recent workers in this group have separated the American members from the Asiatic. The former can be connected through Sibon (= Lepto- gnathus) sibon with the Colubrinae, while the Asiatic genera can- 27—1367 674 THe UNIVERSITY SCIENCE BULLETIN not. The characters of the shields covering the lower jaws serve to distinguish them at once from all other snakes. Nevertheless, the two groups are closely allied to one another, and probably had a common origin. The mouth is peculiar in that the commisure ex- tends far back beyond the fringe of the buccal membrane, while the short, high head and large eye bear a remarkable resemblance to a foetal snake. Another feature of the Dipsadinae is the enormous development of the nasal gland.” Since the oldest applicable generic designation is Dipsas it should be used for the family name, and the Oriental subfamily name would be Pareatinae. Subfamily PAREATINAE This subfamily includes two genera, Pareas and Aplopeltura, which may be distinguished by the following key: Key TO ORIENTAL GENERA OF THE PAREATINAE 1. ‘Scales:in: l5:rows; subeaudals paired: .. =. 72,024.42 eee Pareas 2. Scales in 13 rows, subcaudals single.........................-. A plopeltura Genus Pareas Wagler Amblycephalus (not of Zeder 1803) Kuhl, Isis, 1822, p. 474 nomen nudum; Boie, Isis, 1827, p. 519; (type of genus, Amblycephalus laevis). Pareas Wagler, Natiirliches System der Amphibien . . . 1880, p. 181 (type Pareas carinatus); Theobald, Descriptive catalogue of the reptiles of British India, 1876, p. 203. Eberhardtia Angel, Bull. Mus. Hist. Nat. Paris, vol. 26, p. 291 (type of genus Eberhardtia tonkinensis = Pareas hamptoni (fide Malcolm Smith). Diagnosis: Head very short, thick, distinct from neck; snout blunt; eye large or very large with a vertical pupil; body, especially in arboreal forms, much compressed, tail becoming very slender be- hind vent; scales in 15 rows throughout body; subcaudals paired; ventrals not angulate. Maxillary short, thin, expanded vertically, with four to nine subequal teeth, the front part lacking teeth; anterior mandibular teeth large, decreasing in size posteriorly; prefrontal bone with a backward prolongation more or less com- pletely roofing orbit; no mental groove, anal single. The genus occurs in southeastern Asia, the Malay Peninsula, and Archipelago. There are six species recorded from Thailand, three of which also occur in Malaya. The other three are continental forms that are known either in Burma or Indo-China. SERPENTS OF THAILAND 675 Key TO SPECIES OF Pareas IN THAILAND 1. Vertebral scales not enlarged; body not strongly compressed; labials sepa- rated from orbit by subocular; terrestrial............................ 2 Vertebral scales of median row (sometimes adjoining rows also) more or less enlarged; body strongly compressed arboreal.............0....... 3 2. Scales smooth, usually with, or without an interrupted neck-band; ventrals JESSEMUIMEROUS ey we ee te ee eS Ss margaritophorus Scales keeled, usually with an irregular neck-band; ventrals more numer- OUISH ae eee eer ding aa ee May ener nected had Rae Se CECE maculartus 3. Labials separated from eye by one or more suboculars................ 4 Labials not separated from eye, two or three border orbit............. 5) 4. A single elongate subocular (sometimes fused to postocular)....hamptont IWIO HO) Snedaits SUNN EVES, o666000565600,50000g serdcodcuuucuos carinatus 5. Six supralabials, third and fourth (sometimes fifth also) border orbit; third infralabials forming a median suture with each other.............. laevis Seven supralabials, third and fourth border orbit; second or third infra- labial joining its fellow to form a median suture........... .malaccanus Pareas margaritophorus (Jan) (Fig. 12) Leptognathus margaritophorus Jan, Nouv. Arch. Mus. Hist. Nat. Paris, vol. 2, 1866, p. 8 (type-locality, Siam); Morice, sur la faune de la Cochinchine Francaise, 1875, p. 60; Bocourt, Nouv. Arch. Mus., Paris, 1886, p. 9. Pareas margaritophorus: Theobald, Catalogue of the reptiles of British India, 1876, p. 203; Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, 1885, pp. 1-104; M. Smith, The fauna of British India, es and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1948, p. 117. Amblycephalus margaritophorus: Boulenger, Catalogue of the snakes in the British Museum, vol. 3, 1896, p. 445; Flower, Proc. Zool. Soc. London, 1899, p. 694; Mocquard, Rev. Coloniale, 1907, p. 48; Boulenger, Ann. Mag. Nat. Hist., ser. 8, vol. 14, 1914, p. 485 (Key); M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 3, 1915, p. 215; Werner, Arch. Naturg. Band 87a, Heft 7, 1922, p. 194; Bourret, Invent. Gen. Indochine, 1927, p. 238; Les serpents de l'Indochine, vol. 2, 1936, p. 434. Pareas moellendorffi Boettger, Ber. Offenb. Ver., 1885, p. 125; ibid, 1888, p. 84, pl. 2, fig. 1 (type-locality, Lo-fou-shan Mts., Canton); Cochran, Proc. U.S. Nat. Mus., vol. 77, 1980, p. 37. Amblycephalus moellendorffi: Boulenger, Catalogue of the snakes of the British Museum, vol. 3, 1896, pp. 443-444; A vertebrate fauna of the Malay Penin- sula_. . . Reptilia and Batrachia, 1912, pp. 210-211; Laidlaw, Proc. Zool. Soc. London, 1901, 2, p. 581; Boulenger, Fasciculi Malayenses, Zoology, vol. 1, 1903, p. 170; Smith, Journ. Nat. Hist. Soc. Siam., vol. 1, no. 3, 1915, p. 215 (moellendorffi) and Journ. Siam Soc., Nat. Hist. Suppl., vol. 6, Oct. 1923, p. 204; Wall, Rec. Ind. Mus., vol. 24, 1922, p. 23; Journ. Bombay Nat. Hist. Soc., vol. 30, 1925, p. 245; Smith, Bull. Raffles. Mus., no. 3, 1930, p. 88 (“very common at Bangnara, Patani according to Mr. C. J. Agaard”); Pope, Reptiles of China, 1935, p. 373; Bourret, Les serpents de l’Indochine, 1936, pt. 2, p. 433; Tweedie, Bull. Raffles Mus., Singapore, no. 23, 1950, pp. 191-192. Diagnosis: Prefrontal not excluded from eye; dorsal scales smooth. Loreal separated from the eye by one or two preoculars; eye sepa- rated from labials by a subocular; ventrals, 138-159. THe UNIVERSITY SCIENCE BULLETIN Description of species (from No. 34422): Body somewhat com- pressed; head thickened, distinct from neck, snout blunt, truncate; depression about and behind nostrils; eye large, pupil vertical; area in front of eye swollen; rostral wider than high, excavated below, narrowly visible above; internasals wider than long, narrowly in con- tact laterally with loreal; prefrontals double size of internasals, Fic. 12. Pareas margaritophorus Jan. EHT-HMS No. 13562, Bukit Besar (Mt.), (near waterfall), Na Pradoo, Pattani, province, southem Thai- land. Actual total length, 290 mm.; tail, 48 mm. SERPENTS OF THAILAND 677 entering orbit above preocular; frontal subtriangular about as wide as long, slightly shorter than its distance from tip of snout, much shorter than parietals; supraoculars small, subtriangular, side border- ing orbit slightly curved; parietals much longer than wide, as long as their distance to tip of snout, much narrowed posteriorly. Tem- porals, three, followed by three, all narrow elongate; two upper scales in first series touch suborbital, the lower pushed back behind sixth labial; nasal scale large, subquadrangular, the nostril pierced in the posterior part; loreal longer than high; preocular triangular, not touching supraocular; “suborbital” a narrow curving scale ex- tending from preocular, below and behind eye where it is in con- tact with supraocular; seven supralabials, second to fifth in contact with suborbital; seven infralabials; four bordering first very large chinshields; mental separating first labials and in contact with chin- shields; second pair of chinshields make a narrow median suture; third pair triangular, in contact at a narrow point; no groove be- tween chinshields. Scales entirely smooth, in 15, 15, 15 rows (19 across occiput); ven- trals, 146 not angulate; anal single, subcaudals, 51. The tail sud- denly becomes very slender a short distance behind vent. Color: Head variegated lavender-brown, lighter on labials; a pair of black spots behind parietals and smaller less distinct spots at mouth angle preceded by a white area; rostral and dorsal head scutes darker; a cream-white ring about neck (interrupted dorsally on left side) about three scales wide, and followed by a dark band; body generally light brown but darker on anterior fourth of body and neck; a series of very narrow transverse bars, one scale wide, consisting of scales marked in front with cream, and behind with black; laterally they may widen, and rarely divide; venter gray- white, irregularly spotted or flecked with black, the spots not con- fined to outer part of ventrals, the subcaudal area most strongly marked; chin and neck dirty cream. Measurements in mm: Total length, 320; tail, 69; head-width, 7.1; head-length, 13. Variation: The older specimens, according to Theobold (1868) tend to lose the transverse markings and become a uniform ochra- ceous-brown. It is significant that the scale that I have called the “suborbital” is divided, in the material studied by Blyth and Theo- bald, into “five, six, or seven small shields which exclude the labials from the orbit.” It is possible that his reference is to another species. 678 THE UNIVERSITY SCIENCE BULLETIN In the young specimens the neck-band is mottled with claret-red or orange. The length of one adult is approximately 600 mm. of which 112 mm. is the approximate length of the tail. Two specimens were acquired in 1961 from two widely separated localities, EHT-HMS, No. 1356 ¢ (figured) from Bukit Besar (near waterfall), Pattani province, and EHT-HMS No. 3923 9 , 10 km. N. Chiang Dao, Chiang Mai province. These display the following differences: No. 1356 has three temporals on each side, the two upper touch- ing the large subocular (which is fused to the postocular), the lower one is separated from it and is divided vertically into two scales. The first pair of infralabials form a strong suture behind the mental. The reddish-orange nuchal ring, from two to three scales in width, is complete, followed by a dark band. The ventrals are 149; subcaudals, 41. The venter is gray with numerous scattered gray dots. No. 3923 has the temporals arranged two over one (elongate), over two. There are two touching the subocular. The first in- fralabials do not touch behind mental. On one side there are two loreals. In this specimen the collar is incomplete, consisting of an orange spot covering part of eight scales with two median black dots. Ventrals, 153; subcaudals, 39. Otherwise the two specimens agree in color and markings save that No. 3923 has the venter yellow-ivory with many ventrals bearing a small subquadrangular spot on outer ends and only a few scattered punctate dots. Distribution: The species is known in Thailand from Pattani, Narathiwat, and Chiang Mai. Elsewhere it is reported from Burma, South China, Indo-China and the northern part of Malaya. Remarks: The specimen from Bukit Besar was found in a road- side drain, under a small mass of wet leaves. I suspect that sub- species will be designated for some of the variants of this species but much more material is desirable before this is undertaken. Pareas macularius Blyth (Fig. 18) Pareas macularius Blyth, in Theobald, Journ. Linn. Soc., vol. 10, Zoology, no. 41, 1868, pp. 54-55 (type-locality, Martaban, S. Burma); M. Smith, Rec. Ind. Mus., vol. 42, 1940, p. 480; The fauna of British India, Ceylon, and Burma, . . . Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, p. 118; Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, pp. 1126-1128, fig. 21. SERPENTS OF THAILAND 679 Amblycephalus macularius: Boulenger, The fauna of British India, 1890, pp. 416-417; Catalogue of the snakes in the British Museum, vol. 3, 1896, pp. 444-445; Wall, Rec. Ind. Mus., vol. 25, 1922, p. 24; Journ. Bombay Nat. Hist. Soc., vol. 30, 1925, p. 245; vol. 31, 1926, p. 566. Pareas andersonii Boulenger, Ann. Mus. Civ. Genova, ser. 2, vol. 6, 1888, p. 601, pl. 5, fig. 3 (type-locality, Bhamo and Kakhyen Hills, Burma; fide Smith); separate of same, p. 9. Pareas tamdaoensis Bourret, Bull. Gen. Inst. Publ. Hanoi, vol. 10, 1935, p. 11 (type-locality, Tam Dao, Tong-King, fide Smith); Les serpents de I’Indo- chine, vol. 2, 1936, p. 431. Diagnosis: Resembles Pareas margaritophorus in color and mark- ings but differs from it however in having a more compressed body. There are keels on five or seven median scalerows, and a somewhat larger series of ventrals and subcaudals. Description of species (from EHT-HMS No. 31798 9° , Thailand): Rostral large, slightly wider than high, its upper edge, as seen from above, a straight line; internasals more than twice as wide as long, notching upper edge of nasal, narrower than prefrontals; latter Fic. 13. Pareas macularius Theobald. EHT-HMS No. 3179892, Thailand. Actual total length, 210 mm.; tail, 49 mm. 680 THe UNIVERSITY SCIENCE BULLETIN wider than long, touching preocular and loreal, entering orbit; frontal short, six-sided, its length a little greater than width, slightly longer than its distance from tip of snout, shorter than parietals; latter scales longer than their distance from tip of snout; nostril in a single nasal, somewhat triangular; loreal about as wide as high; one preocular, one postocular; a large narrow curving subocular; supraocular not longer than eye; temporals, 2 + 3, but irregular in second (and third) rows; seven supralabials, four or three touching subocular, last much the longest; mental small, touching first pair of chinshields, separating first labials from one another; eight (seven) infralabials, four (five) touch first chinshields which are longer than wide; second and third pairs transversely widened; no groove under chin; first ventral touching third pair much larger than following ventrals; eye higher than its distance from mouth; 20 scales across back of head; 15 rows at neck, middle of body, and before vent; ventrals, 163; subcaudals, 55 paired; anal single; mandibular teeth, 29, the anterior five times as large as the posterior teeth. Color: Top of head heavily speckled with brown; a pair of black spots behind parietals; diagonal brown line from upper level of eye to mouth angle; on sixth and seventh labials a suggestion of a second line; brown flecks on supralabials; a somewhat zigzag cream nuchal band crosses occiput and nape and merges with light color of chin and throat behind jaw angle; cream band followed by a W-shaped black mark; general body color brown with approxi- mately 60 transverse rows of scales with black dots, each dot bordered anteriorly by cream, the rows irregular and often incom- plete. Outer part of ventrals and subcaudal region flecked with brown. Measurements in mm.: Total length, 210; tail, 49. Pareas hamptoni Boulenger Amblycephalus hamptoni Boulenger, Journ. Bombay Nat. Hist. Soc., vol. 11, 1905, p. 236, pl—(type-locality, Mogok, Burma); Ann. Mag. Nat. Hist., ser. 8, vol. 14, 1914, p. 485; Parker, Ann. Mag. Nat. Hist., ser. 9, vol. 15, 1925, p. 305; Wall, Rec. Ind. Mus., vol. 24, 1922, p. 26; Journ. Bombay Nat. Hist. Soc., vol. 30, 1925, p. 245. Pareas hamptoni: M. Smith, Ann. Mag. Nat. Hist., ser. 10, vol. 6, 1930, p. 681; Rec. Ind. Mus., vol. 42, 1940, p. 480; The fauna of British India, Ceylon and Burma ... . Reptilia and Amphibia, Dec. 1948, p. 120 (N. E. Siam, near Pak Lai, lat. 18° upper Mekong, etc.) Eberhardtia tonkinensis Angel, Bull. Mus. Hist. Nat. Paris, vol. 26, 1920, p. 20) Hes LS (type-locality, Laokay, Hong-King), and ibid., ser. 2, vol. 1, 29, p. 80. SERPENTS OF THAILAND 681 Amblycephalus tonkinensis: Bourret, Les serpents de Indochine, vol. 2, 1936. p. 428, fig.; Pope, The reptiles of China, 1935, p. 378, fig.; Amblycephalus carinatus hainanus M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 6, 1923, p. 204 (type-locality, Hainan ). Amblycephalus carinatus berdmorei: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 88 (part.). Diagnosis: A small species, vertebral scales enlarged; body strongly compressed; frontal shorter than parietal; prefrontal enter- ing eye. Head sharply distinct from neck; one preocular; loreal excluded from eye, or barely touching it. Description of species: Snout short, blunt; rostral a little higher than wide; internasals about half as long as prefrontals; prefrontals entering orbit; frontal longer than its distance from end of snout, shorter than parietals; loreal as high as long, or a little higher, not or scarcely touching eye; one preocular; one postocular; a long crescentic bandlike subocular connecting pre- and postocular or postocular also united with subocular; temporals, 2+ 2 or 2+ 3; seven or eight supralabials, fourth and fifth below middle of eye; scales in 15 rows, scales smooth or median rows slightly keeled; vertebral series, and sometime adjoining rows feebly enlarged: ventrals, ¢ 191-196; subcaudals, 93-98; of female, ventrals 180-194, subcaudals 73-87. Color: Similar to monticola. Brown above with vertical blackish bars on sides or sometimes extending across back; a black line from above eye to nape, and another from behind eye to angle of mouth; top of head more or less thickly spotted with black; yellowish be- low, dotted brown. Remarks: The species is seemingly rare in Thailand. Malcolm Smith’s records it from “near Pak Lai lat. 18° upper Mekong,” N. E. Thailand, and under the name Amblycephalus carinatus berdmorei, he reports the species as being common at “Bangnara” (= Narathi- wat, Narathiwat), Thailand. Elsewhere it is known in northern Burma, Tong-King, Annam, and Hainan. Pareas carinatus Boie (Fig. 14) Amblycephalus carinatus Boie, Isis, 1828, p. 1035 (type-locality, Java): Boulenger, Catalogue of the snakes in the British Museum, vol. 3, 1896, pp. 445-446; Barbour, Mem. Mus. Comp. Zool. Harvard Col., vol. 44, 1912, p. 1388 (‘ ‘Bangnara, Patani” ); Smith, Journ. Nat. Hist. Soc. Siam, vol. oF 1916, pp. 163-164; Wall, Rec. Ind. Mus., vols 24.1990 p25 -me Bombay Nat. Hist. Soc., vol. 30, 1925, p. 426; Bourret, Les Seen de l’Indochine, pt. 2, 1936, p. 435, fig. 682 Tue UNIVERSITY SCIENCE BULLETIN Pareas berdmorei Theobald, Catalogue of the reptiles in the Asiatic Society’s museum, 1868, p. 63 (type-locality, Tenasserium ). Pareas carinatus: Cochran, Proc. U. S. Nat. Mus., vol. 77, 1930, p. 37 (Bang Suk, near Pak Jong, Thailand); M. Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, p- 121; Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, pp. 1128, 1130, fig. 22 (Loei province). Amblycephalus carinatus berdmorei: Smith, Bull. Raffles Mus., no. 3, 1930, p. 88 (in part.). Diagnosis: Body strongly compressed; vertebral scales somewhat enlarged or not; frontal as long or longer than parietal; prefrontal does not border orbit; head distinct from neck; separated from eye by from two to eight suboculars. Description of species (from EHT-HMS No. 3626, 10 km. N. Chiang Dao): Body strongly compressed, tail prehensile; head large, distinct from neck; upper lips flaring out. Rostral higher than wide (2% x 3% mm.), not, or but scarcely visible from above; internasals triangular, much wider than long, their mutual suture longer than that between prefrontals and forming a suture with the loreal nearly as long as that with rostral; prefrontals a half wider than long, touching loreal and preocular laterally, narrowly separated from eye; frontal with the sides nearly parallel, much wider than supraoculars, longer than its distance to end of snout, slightly shorter than parietal; nasal single, longer than high, nostril pierced in posterior part; loreal higher than long, touching second supralabial and two preoculars of which upper one is highest but widely separated from frontal; one postocular; supraocular touches the postocular and upper anterior temporal; parietals scarcely larger in area than frontal, bordered laterally by four temporals and posteriorly by five postparietals; two or three suboculars sep- arating supralabials from orbit; supralabials, 8-7, first four higher than wide, last two longer than high, the last about three times as long as wide; infralabials, 9-8, first forming a postmental suture, first four bordering first chinshields and first supralabial on left side; mental as wide as rostral; three unequal pairs of chinshields followed by a large azygous scale much larger than first ventral; eye large, its diameter greater than its distance from anterior part of snout. Three anterior temporals, four secondary temporals, some- what irregular. Scalerows: 20 (occiput), 15, 15, 15, the median dorsal row (vertebrals ) not or scarcely enlarged; ventrals, 174; anal undivided; subcaudals, 72, all save the ultimate divided; several median scale- rows keeled. : SERPENTS OF THAILAND 683 Fic. 14. Pareas carinatus (Boie). No. 3626, 10 km. north of Chiang Dow, Chiang Mai province, northern Thailand. Actual total length, 488 mm.; tail, 108 mm. 684 THe UNIVERSITY SCIENCE BULLETIN Color: Head light brown on top and in temporal region, with some black dots. Rostral, supralabials, chin, ventrals, subcaudals and outer scalerow yellowish ivory a little more yellow anteriorly. A more or less connected black line arises back of eye passes back along temporal region to side of neck, narrowly separated from its fellow from opposite side; a hair-fine black line from lower part of eye to lip, crossing subocular, lower anterior temporal and the last two supralabials; on dorsum generally light or olive brown with about 71 transverse darker marks crossing body, these rather in- distinct, and usually the width of one scalerow, formed by separate blackish marks on scales and somewhat more scattered pigment; outer edge of ventrals with some fine dots or flecks of dark pigment and a more or less definite row of dark dots; transverse dark marks dim on tail, the subcaudal dots more scattered. Measurements in mm.: Total length, 488; tail, 108; width of head, 11; length of head, 16; depth of head, greatest, 9. Variation: A specimen from Loei province has 77-78 bands ( dif- ferent on two sides) while those on a female specimen from the same province are so dim they cannot be counted. The temporals may have a fairly regular arrangement of 3+ 3-+ 4 or 3+4-4 4. There are five small ocular scales in both of these specimens while there are six in the described form on one side. The ventrals in five specimens examined (one unnumbered, from an uncertain Thai locality) vary from 158 to 181; the subcaudals from 68 to 72 (Malcolm Smith reports 170-184 for ventrals, sub- caudals 60-80). Data on Pareas carinatus NUMBER 31796 31797 3626 TSO Calitiy nese cet verse cone rs eet bee Loei Loei Chiang Mai Motallenathie se see ey 430 522 488 of Dell eas aera arent ac ey et a cae 41+ 111 108 Widtheotzhesdies te ee ne ee ae 10 11.2 11 Bengthyotenead eercw rs eee ee 16.2 19 16 Sealerows at middle............... 15 15 15 Went Sec ae a ee 181 178 174 Sulbcawdalsivcse aon co aaa 26+ 71 72 iPreoculars ats se ee ane 1-1 1-1 2-2 IPostoculans oe en eee ee ee 1-1 1-1 1-1 Subocularsis= se. oe oe ee 3-3 3-3 3-3 Supralabials\ yee essere eee =e =H 8-7 Uniralabial siete ines wee ea ears 8-9 8-5 9-8 SERPENTS OF THAILAND 685 Distribution: The species is widely distributed in Thailand, speci- mens having been taken in the provinces of Chiang Mai (Chiang Dao), Loei, Narathiwat, and Nakhon Si Thammarat (Khao Luang Mt.). Outside of Thailand it is known in Burma, the Indo-Chinese region from Lat. 19° N, south throughout Malaya and the Indo- Australian Archipelago. Certain subspecies have been described that may be worthy of retention. Pareas laevis Boie in Boie Pareas laevis Boie, in Boie, Isis, 1827, p. 520 (type-locality, Java); Ginther, The reptiles of British India, 1864, p. 328; Tweedie, The Snakes of Malaya, 1954, p. 54, (Northern Malaya). Amblycephalus laevis: Boulenger, The fauna of British India, Ceylon and Burma; Reptilia and Batrachia, 1890, p. 415; Catalogue of the Snakes in the British Museum, 2nd Ed., vol. 3, 1896, pp. 441-442; A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 209; Laidlaw, Proc. Zool. Soc. London, 1901, 2, p. 581; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia 1917, p. 276; M. Smith, Bull. Raffles Mus., no. 3, Apr. 1930, p. 88 (“Khao Ram in the Nakhon Si Thammarat mountains” ). Diagnosis: Small snakes reaching a length of 545 mm., the body strongly compressed; no preocular; loreal and prefrontal entering eye; two or three labials bordering orbit; first chinshield undivided, in contact with mental; third infralabial large touching its fellow; two pairs of large chinshields broader than long; brown with ir- regular dark crossbands; sides orange, venter brown or yellowish, spotted with brown or with transverse brown spots on sides. Description of species: Rostral broader than high; prefrontals entering eye; frontal as long as broad or a little broader than long, as long as, or a little longer than its distance from end of snout, two thirds as long as parietals; supraocular very small; loreal elongate bordering orbit; no preocular, one or two postoculars; temporals, 2+ 2; six supralabials third and fourth (third, fourth and fifth) bordering orbit, sixth very long; mental in contact with first un- divided chinshield; third infralabial very large forming a suture with its fellow; a small azygos chinshield followed by two pairs of large chinshields, broader than long. Scales in 15 rows throughout, smooth, vertebral row enlarged; ventrals, 148-176, anal, single, subcaudals, 34-59. Eye large; no longitudinal groove under chin. Color: Brown above with irregular transverse blackish cross- bands; sides dull orange; lower parts brownish or yellowish, spotted with brown or with transverse brown spots on sides. 686 Tue UNIVERSITY SCIENCE BULLETIN Measurements in mm.: Total length, 545; tail, 65. Remarks: The species has been collected only once in Thailand. Dr. Malcolm Smith found it on Khao Ram in Nakhon Si Tham- marat. The species, while having a distribution in Sumatra, Java and Borneo, appears to be absent in the lower part of the Malay Peninsula. (Data chiefly from Boulenger, 1912, and Smith, 1930.) Pareas malaccanus (Peters ) Asthenodipsas malaccana Peters, Monatsb. Akad. Wiss. Berlin, 1864 p. 27, pl. —, fig. 3. (type-locality, “Malacca” ). Amblycephalus malaccanus: Boulenger, Catalogue of the snakes in the British Mus., 2nd Ed., vol. 3, 1906, pp. 442-443; A vertebrate fauna of the Malay Peninsula. . . Reptilia and Batrachia, 1912, pp. 209-210; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, pp. 276-277 (Sumatra, Borneo, Mentawei Islands); M. Smith, Bull, Raffles Mus., no. 3, Apr. 1930, p. 88 (Bhetong, “Patani” = Yala province). Pareas malaccanus: Tweedie, The Snakes of Malaya 1954, p. 36. Diagnosis: Small snakes, reaching half a meter in length; no pre- ocular; loreal and prefrontal bordering eye; two postoculars; third and fourth labials bordering orbit; second and third infralabials very large in contact mesially with each other; yellowish or pale brown above with rather irregular dark-brown cross-bars interrupted mesially. Description of species: Rostral a little broader than deep; frontal as long as broad or a little broader than long, as long as or a little longer than its distance from tip of snout, much shorter than the combined parietals; supraocular small; prefrontal and loreal enter- ing orbit; no preocular; two postoculars; temporals, 2 + 2 the two upper ones sometimes fused; seven upper labials, third and fourth bordering orbit, seventh enlarged; mental in contact with a small azygous scale; second and third infralabials very large, usually in contact mesially on chin; two pairs of large chinshields, broader than long; no mental groove. Scales in 15 rows, smooth, the three median rows a little en- larged; ventrals, 154-170; anal, single; subcaudals, 26-55. Color: Yellowish or pale brown above, with rather irregular dark-brown cross-bars which are interrupted on back; vertebral scales yellowish; head sometimes whitish; sides and lower surface of neck black; belly whitish, uniform or speckled with dark brown. Measurements in mm.: Total length, 440; tail, 50. Distribution: The species has been taken in Thailand only at Bhetong in Yala province. It also occurs in Malaya, Sumatra and Borneo. SERPENTS OF THAILAND 687 Remarks: Malcolm Smith states that the Bhetong specimen had only the median dorsal scalerow enlarged. Boulenger has sug- gested that the type-locality should be “Malay Peninsula,” rather than “Malacca.” Genus Aplopeltura Duméril and Bibron Aplopeltura Duméril and Bibron, Mem. Acad. Sci. Paris, vol. 23, 1853, p. 463 (type of genus, Amblycephalus boa Boie). Haplopeltura Boulenger, Catalogue of the snakes in the British Museum, 2nd Ed., vol. 3, 1896, p. 489 (emendation). Diagnosis: Head strongly distinct from neck; eye large with vertical pupil; nasal undivided; body strongly compressed with long tail. Scales smooth in 13 rows without pits; vertebrals en- larged; costal scales placed somewhat obliquely; subcaudals single; no mental groove. Maxillary short, thin, expanded vertically, with five subequal teeth preceded by a short space without teeth; 12-13 mandibular teeth, decreasing in size posteriorly; skull with a wide vacuity be- tween parietal, frontal and sphenoid bones. Only a single species is recognized. The genus is widely dis- tributed from the Philippines throughout the Malay Peninsula and Archipelago. Aplopeltura boa (Boie) (Fig. 15) Amblycephalus boa Boie, Isis, 1828, p. 1034 (type-locality, Java); Giinther, Catalogue of colubrine snakes, 1858, p. 184; The reptiles of British India, 1864, p. 325. Dipsas boa: Schlegel, Essai sur le physionomie des Serpents, vol. 2, 1837, p. 284, pl. 11, figs. 29, 30; Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 925, pl. 40, fig. 3. Aplopeltura boa: Duméril, Bibron and Duméril Erpétologie générale, vol. 7, 1854, p. 444; Smith, Bull. Raffles Mus., no. 3, 1930, p. 88 (“hills to the east of Chumporn = [Chumphon] . . . 1000 ft. elev.”). Haplopeltura boa: Boulenger, Catalogue of the snakes in the British Museum, Qnd Ed., vol. 3, 1896, pp. 439-440; Flower, Proc. Zool. Soc. London, 1899, p. 693; Boulenger, Fasciculi Malayenses, Zool., vol. 1, 1903, p. 170; A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 208; Taylor, Dept. Agri. Nat. Res. Bureau of Science, Pub. 16, Feb. 11, 1922, pp. 280-283, pl. 34, figs. 7-9 (head); Smith, The fauna of British India, Ceylon, and Burma . . . Reptilia and Amphibia, vol. 3, Ser- pentes, 1943, p. 122 (Nakhon Si Thammarat, Chumphon); Tweedie, The snakes of Malaya, 1954, p. 36-37, fig. 6c (head). Diagnosis: Slender compressed snakes reaching a length of 750 mm.; snout short, head thickened; two (or three) enlarged occipital scales; two or three loreals; a preocular, several suboculars and a postocular; two pairs of infralabials meet behind mental. Scales in 13 rows, vertebrals enlarged; subcaudals single. The color var- 688 THe UNIVERSITY SCIENCE BULLETIN iable from yellowish-gray to brown, blotched, mottled, or clouded with lichen-green or brownish red; below yellowish, mottled with brown. Description of species (from P.K. No. 101, Na Pradoo, Pattani [near Bukit Besar] Thailand: Rostral narrowly visible above, some- what more than one-and-one-half times as high as wide, with a high-arched notch at its base, above which the scale is depressed or excavated; internasals extended transversely, suture between scales about one third as long as transverse length of scale (left in- ternasal abnormally broken, the severed part having the appearance of a loreal); prefrontals large, their common suture little less than half width of one scale; frontal hexagonal, sides nearly parallel, anterior and posterior parts having same shape but anterior slightly Fic. 15. Aplopeltura boa (Boie). P.K. No. 101, Na Pradoo, Pattani province, southern Thailand. Actual total length, 719 mm.; tail, 223 mm. SERPENTS OF THAILAND 689 wider; length of frontal equal to or a little greater than its distance from tip of snout and equally as long as parietals; latter scales relatively small with an elevation near middle; pair of enlarged occipital scales bordering parietals. Nostril pierced in a single triangular nasal; three loreals of different sizes, upper largest; two distinct preoculars, two post- oculars, upper largest; three suboculars; supraocular large bending outward behind eye, with a slight curved emargination on outer edge immediately above eye; three rows of irregular temporals, upper anterior largest, bordering upper postocular; nine (or ten) supralabials, separated from eye by suboculars. Mental small with two pairs of infralabials in contact with each other behind mental; eleven infralabials, posterior one largest; no mental groove present; chinshields followed by two azygous scales, and these followed by three pairs of chinshields which are large but somewhat irregular. Scale-formula: 26 (bordering occiput), 13, 13, 13; median scale series enlarged; lateral scales narrow, elongate, somewhat oblique; ventrals, 165%; anal single; subcaudals, 108, all single. Eye large, its diameter equal to its distance from snout-tip; pupil vertical; maxillary teeth five, preceded by a short vacant space, the teeth subequal; 12 (13) mandibular teeth, anterior ones very large, fanglike, diminishing posteriorly in size. Tongue with longitudinal black lines. Head large, much thickened, sharply set off from body. Color: Above brownish gray with indefinite dark markings and flecks of black; lower on sides mottled with irregular smoky-black blotches; venter generally lichen-gray with some cream color. Head with a conspicuous cream-white spot covering most of the side of face in front of eye; snout tip with numerous blackish flecks; a large blackish spot covers occipital, supraocular and interorbital areas, bordered laterally, behind orbits, with a yellowish stripe that is bordered below by a broken black band; a vertical black mark from eye to lip. Measurements in mm.: Total length, 719; tail, 223; width of head, 10.5; length of head, 20. Distribution: The species is widespread, extending from Minda- nao, Palawan, and Balabac in the Philippines, through Borneo, Java, Sumatra, Malaya, and Peninsular Thailand. In Thailand specimens are known from Bukit Besar (mt.), Pat- tani, Nakhon Si Thammarat, and Chumphon. The described speci- men is from Na Pradoo at the base of Bukit Besar. 690 THe UNIVERSITY SCIENCE BULLETIN Variation: The known variation reported in the literature for ventrals is 148 to 191; subcaudals 88 to 127. Scale counts for main- land specimens given by Malcolm Smith (1943) is 166-175 for ventrals, 106-122 for subcaudals. There may be three or two pairs of chinshields and eight to ten supralabials. The largest specimen, a female, measures 835 mm. in total length; the tail-length being 265 mm. The wide range with concomitant variation suggests the prob- ability that subspecies occur which merit recognition. Remarks: The single specimen taken in Thailand and reported by Dr. Malcolm Smith (1943) was colored in life as follows: “pale gray with narrow black cross-bars or almost complete bands; top of head and vertebral scales red, the former speckled with black.” A specimen taken in Agusan province, Mindanao, Philippines was discovered in a small lichen-covered dead tree. When it fell to the ground around an accumulation of dead twigs, it remained motion- less in the same position in which it had fallen, and was picked up in this semirigid condition. It was seemingly mimicking the lichen- covered twigs and small branches, since its color resembled markedly that of the twigs. Famity CoLusripaE Cope The family Colubridae contains a large proportion of the living serpents of the world, the bulk of which are without poison and of those with poison-glands and grooved teeth, very few are of danger to man. The family has been broken up into several subfamilies. In this work I have treated the Thai forms in seven subfamily groups. The degree of differentiation in these groups is doubtfully of equal value. They may be differentiated by the following key. Key To THAI SUBFAMILIES OF THE COLUBRIDAE 1. No ventral shields; aquatic snakes (fresh and salt water) with from 100 to 150 small juxtaposed scales around middle of body;........ Acrochordinae Ventral shields widened; hypapophyses present or absent on posterior dorsal vertebrae. 34 Ges a eo ee an eS . Dentary bone detached from articular posteriorly; teeth ungrooved, numerous, 30 to 50 in each maxilla and closely set, equal in size, often sub- spatulate or bayonet-shaped; hypapophyses throughout the vertebral col- UMMM terrestrial er oie tects Mere nae ee ne stairs ete en ele o Sibynophinae Dentary bone not detached from articular posteriorly................. 3 3. Aquatic snakes (fresh and marine waters), maxillary teeth 10-22, some usually enlarged, grooved, fanglike; nostril crescentic on dorsal surface of snout; hypapophyses developed throughout the vertebral column, all (ex- cept Cantoria) proably not exceeding a meter in length... . Homalopsinae Terrestrial serpents. Nostril not crescentic, not situated on dorsal part of Fs}. Ue neh cree Seen Mies tT bance tricnes tee Nr RG ast Ryal eek iia ae reas Ai a ooo aa & 4 bo bo SERPENTS OF THAILAND 691 4, A supraorbital bone; scales completely or almost completely attached to the cutis, and separated from one another by naked skin; head-shields granular-or shields normal; labials usually with everted margins; nostril in a concave shield; maxillary teeth 15-35, none grooved or fanglike, X enoderminae No supraorbital bone; scales normally not separated by naked skin, not attached alone to the cutis; maxillary teeth with or without grooved AeA eS aeP ANG raph temper Ce CNA SNe WininOs hanna Ee Gece se ore Orus an tne Sf Una Sata aaa 5) 5. Maxillary teeth, ten to eighteen, the last two or three teeth grooved and fanglike. Hypapophyses present or absent on posterior dorsal vertebrae; mostly arboreal snakes, none completely aquatic.............. Boiginae Maxillary teeth 6-33 without grooves; some may develop enlarged pos- teriormtan glikesteethim acc anon ee ee eas 6 6. Hypapophyses present throughout the vertebral column (Natriz, Rhab- dophis Pseudoxenodon Macropisthodon Parahelicops, Opisthotropis) Natricinae Hypapophyses absent in posterior vertebrae Gonyosoma, Elaphe, Ptyas, Lycodon, Dinodon, Oligodon, Liopeltis, Dryocalamus, Calamaria, Plagi- opholis, Pseudorhabdion, Ahaetulla.............0000.-+.005- Colubrinae Subfamily XENODERMINAE Cope Xenoderminae Cope, Ann. Rept. National Museum 1898 (1900) p. 731. M. Smith, Ann. Mag. Nat. Hist., ser. 11, vol. 8, Mar. 1939, pp. 393-395; Werner Mitt. naturh Mus. Hamburg, vol. 26, 1909, p. 206); The fauna of British India, Ceylon and Burma, . . . Reptilia and Amphibia, vol. 3, Ser- pentes, 1943, pp. 123. Malcolm Smith, loc. cit. has placed the following genera in this subfamily: Xenodermus, Nothopsis, Stoliczkaia, Fimbrios and Acha- linus, genera, most of which formerly had been considered in the Acrochordinae. Of these only a single genus and species is known in our territory. This is Xenodermus javanicus, known in Thailand from two specimens taken by me in the province of Yala near the La Doo Tin Mine, Bendang Stah. Genus Xenodermus Reinhardt Xenodermus Reinhardt, Overs. Viden. Selsk. Forh., 1836, p. 6 (type of genus, javanicus ). Gonionotus Gray, in Stoke’s, Discovery in Australia, vol. 1, 1846, p. 502, (type, plumbeus). Diagnosis: Body slender, feebly compressed. Scales varied in middorsal region: a median series of somewhat enlarged scales alternating with smaller scales; the two adjoining series equally enlarged, but scales not contiguous in row, separated by two or more very small scales; a dorsolateral series of enlarged, somewhat elongated scales, separated from each other by small scales; scales on flanks small, larger as they approach ventrals, the scales for the most part separated from each other by skin. Maxillary with 15 equal teeth; pupil vertical; internasals and prefrontals present, but remainder of head covered with very small 692 THE UNIVERSITY SCIENCE BULLETIN granular scales bearing keels; tail elongated, with undivided sub- caudals. Zygapophyses expanded laterally. Only a single species is known. It ranges in the Malay Peninsula from southern Tenasserim, Thailand, and Malaya and extending to the islands of Java, Sumatra, and Borneo. It is said to be com- mon at an elevation of from 500-1100 meters altitude in central Java, living in wet earth below the surface of the ground. It feeds chiefly on frogs. The species is ovoviviparous. Only two specimens are known to have been taken in Thailand. Xenodermus javanicus Reinhardt (Fig. 16) Xenodermus javanicus Reinhardt, Overs. Viden. Selsk. Forh., 1836, p. 6, (type- locality, Java); K. Danske Vidensk Selsk. Skrift, vol. 10, 1943, p. 257, pl. 2, figs. 1-8; Boulenger, Catalogue of the snakes in the British Museum, vol. 1, 1893, p. 175; M. Smith, Ann. Mag. Nat. Hist., ser. 11, vol. 3, 1939, p. 393; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, pp. 44-45, figs. 24-25: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 40; Tweedie, The snakes of Malaya, 1954, p. 37, (Penang and Perak, Malaya). : Diagnosis: See diagnosis of genus. Description of species (from No. 1350, La Doo Tin Mine, Benang Star [Bendang Stah], Yala). Rostral subtriangular, not reaching top level of snout, and not visible above, with a keellike elevation along lateral edges; nasal scale single but with an entrant suture from behind; nostril lunate, a deep depression on posterior part of scale, its back edge flaring up; upper edge with a keellike ridge; two nasals meetings mesially on front of snout; a pair of internasal scales lying back of nasal on upper surface of snout, widening mesially and in contact with each other, their posterior edges keeled; these followed by a pair of larger prefrontal scales, narrowly sep- arated anteriorly by a single scale, the two diverging posteriorly, separated by several small scales; 17 or 18 small supralabials; re- mainder of head covered with small keeled juxtaposed tubercular scales, about 20 rows between eyes; several scale series lying above supralabials, those following nasals, and scales in frontal area somewhat larger; about 24 minute scales around eye, the three or four above eye largest; mental very small, broadly triangular; first pair of infralabials in contact behind mental; a pair of dis- tinguishable chinshields separated by a row of granules; 21-22 infralabials, a little better differentiated from other scales than supralabials; 13-14 scales between chinshields and first ventral; 88 tubercular scales around head at first ventral; 58 to 60 scales at middle of body; 44 in front of vent. SERPENTS OF THAILAND Fic. 16. Xenodermus javanicus Reinwardt. No. 1350, La Doo Tin Mine, Yala, southern Thailand. Actual total length, 433 mm.; tail, 282 mm. Three specialized median scalerows, that in the middle composed of alternating large and small scales, the larger ones more or less alternating with a pair of scales somewhat lateral to it; scales of the two bordering rows are not in contact; a few small scales inter- calated among the three rows; separated from these median rows by five or six small granular scalerows is a dorsolateral discontinuous 694 Tue UNIVERSITY SCIENCE BULLETIN row of scales about half as large as the median rows, the scales longer than wide, separated each from the other by from one to three granular scales; below this row seven or eight irregular rows of fine granules, and below this growing a little larger as they ap- proach ventrals are eleven or twelve rows of scales more or less separated by naked skin; anteriorly the scales of the median rows are small, gradually increasing in size for some distance. Most body scales keeled. Ventrals, 172 +1 pair; anal single; subcaudals single, 113; a widely separated pair of scales precedes first subcaudals (the tip of the tail has been injured). (A second specimen, No. 1372 ¢ has the ventrals, 169 + 1 pair; anal single; 124 single subcaudals, posteriorly narrowing, becoming notched then compressed, and keeled with slight terminal mucrones ). Color in life: Head brownish, tip of snout blackish with some whitish marks on nasals and rostral; dorsum brown, light dove-gray on sides; venter gray, each ventral with a transverse gray-brown mark along its anterior edge; subcaudals barred with brown; chin and throat whitish and a whitish mark bordering the anterior ventrals. Measurements in mm. (Nos. 1350 and 1372). Total length, 433, 386; snout to vent, 282, 243; tail, 151, 143; width of head, 7.5, 6.4; length of head, 11.5, 10. Variation: The number of scales round the body varies between 40 (fide Boulenger 1912) and 60 in the specimen described; M. Smith (1930) reports a Burmese specimen with 170 ventrals, 120 subcaudals and 60 scales around body. The specimen is a female. The known variation in ventral counts is from 170 to 185; the subcaudals from 124 to 165. The snake reaches a length of 453 mm. The tail length is 228 mm. Distribution: These two specimens taken at the La Doo mine are the only ones that have been found in Thailand. Elsewhere one specimen has been found in southern Burma, and there are two records for Malaya (Penang and Perak). The species occurs also in Sumatra, Borneo, and Java. Remarks: The species is reported as being found in Java, bur- rowing in cultivated land at an elevation of 3000 ft., always near water. The specimen herewith described was found coiled in a soggy rotten log in a stream; it was discovered when the log was kicked to pieces. It was rolled into a ball in a cavity and it re- mained more or less in a ball after it was picked up and examined. SERPENTS OF THAILAND 695 The other specimen likewise was obtained by kicking to pieces a similar log in the same rivulet, a few hundred feet higher at an elevation of about 2500 ft. Subfamily SrpyNOPHINAE Dunn Remarks: The subfamily of the Sibynophinae consists of three genera, Sibynodontophis, confined to Centeral America, Parasiby- nophis confined to Madagascar, and Sibynophis to the Oriental Region. Snakes of this group appear to effect autotomy when the tail is grasped by a predator. This would account for the fact that a large proportion of captured specimens have incomplete tails. (See Taylor and Elbel, 1958. ) Genus Sibynophis Fitzinger Sibynophis Fitzinger, Systema reptilium, 1843, p. 26 (type Herpetodryas gemminatus ). Polyodontophis Boulenger, The fauna of British India including Ceylon and Burma; Reptilia and Batrachia, 1890, p. 301. (Substitute name for Enicognathus preoccupied. ) Diagnosis: Dentary loosely attached, and completely detached posteriorly from end of articular. Teeth very numerous on maxil- lary, palatine, pterygoid, and dentary, closely set, of approximately the same size; head scarcely distinct from neck; pupil round. Body cylindrical, the ventrals without lateral keels or notches; paired sub- caudals; hypapophyses developed throughout vertebrae; anal di- vided. Oviparous. Three species occur in Thailand: Sibynophis melanocephalus Gray, a southern, peninsular form; S. collaris (Gray) seemingly widely distributed; and S. triangularis Taylor and Elbel which ranges in southeastern Thailand, Indo-China, and north into Assam. Key To Species oF Sibynophis In THAILAND 1. Tenth supralabial much the largest; black triangular nuchal spot, bordered laterally by a cream or yellow stripe continuous with cream lines on la- Rial see eats ices coe ue coh aes nest ata cloned NiNin pelle ance eae ag ee Nee triangularis Tenth supralabial not largest, nuchal dark band not triangular......... 2. Ten supralabials, eighth much the largest; two postoculars, only upper bo touches parietal; temporals, 1 + —— collaris Nine supralabials; temporals, 2 + 2; two preoculars, both touch parietal; ventrals, 144-169; subcaudals, 120-145; brown or blackish above with two lighter yellowish or bluish-gray longitudinal streaks, each of which may bearaseries\ ols blacks spotsi: isi. aegis tenn. eel melanocephalus 696 ‘THe UNIVERSITY SCIENCE BULLETIN Sibynophis melanocephalus (Gray ) (Fig. 17) Lycodon melanocephalus Gray, Ilustrations of Indian zoology, vol. 2, 1834, pl. 83, fig. 2, (type-locality, Singapore? ). Herpetodryas prionotus Cantor, Proc. Zool. Soc. London, 1893, p. 52, (type- locality “Malacca” ). Ablabes melanocephalus: Giinther, Catalogue of the colubrine snakes in the collection of the British Museum, 1858; The reptiles of British India, 1864, p. 229. Polyodontophis geminatus: (part.), Boulenger, Catalogue of the snakes of the British Museum (Natural History), vol. 1, 1893, pp. 185-186; A vertebrate fauna of the Malay Peninsula from the Isthmus of Kra to Singapore; Reptilia and Batrachia, 1912, pp. 120-121 (part.); de Rooij, The reptiles of the Indo- Australian Archipelago, vol. 2, 1917, pp. 54-55 (part.); M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 158 (“Bangnara, Patani”). Sibynophis melanocephalus: Tweedie, The snakes of Malaya, 1954, pp. 63-64; Batchelor, Malay Nat., vol. 12, pp. 103-112. PSibynophis geminatus insularis: Mertens, Senckenbergiana, vol. 9, 1927, pp. 240-241, (type-locality Sabana, North Sumatra). Sibynophis geminatus: Sworder, Singapore Nat., 1922, p. 60 (Singapore); M. Smith, Bull Raffles Mus., no. 3, Apr. 1930, p. 40 (Narathiwat, Thailand). Diagnosis: A median dark stripe flanked by two lighter stripes that bear small transverse spots or bars; ventral scales with an in- flux of lateral color, and a series of black spots on outer part of ven- trals. Scales in 17 rows; nine supralabials, fourth and fifth touching eye. Description of species (from P.K. No. 106 *, Na Pradoo, Pattani, southern Thailand): Rostral more than twice as broad as high, visible above as a narrow line, the outer lower edges of scale some- what free and tending to overhang mental and two adjoining in- fralabials; internasals broader than long, more than half size of prefrontals, touching both nasals; prefrontals much wider than long, touching posterior nasal, loreal, and preocular laterally; frontal longer than wide, its length equal to its distance from posterior edge of parietals, greater than its distance from end of snout; parietals elongate, bordered behind by five body-scales, laterally by two temporals, posterior much elongate; nostril between two nasal scales, of which posterior is larger and higher than anterior; loreal small, a little higher than wide; preocular larger than loreal, barely reaching upper surface of head; supraocular large, as long as frontal but less wide; two postoculars, both touching parietal; temporals 2-2, lower anterior largest; both anterior temporals in contact with lower postocular. * Collection of Nai Prayoon Kananuracks. SERPENTS OF THAILAND Fic. 17. Sibynophis melanocephalus (Gray). P.K. No. 106, Na Pradoo, Pattani; actual length, 288 mm.; tail, 58 + mm. (part missing). Nine supralabials (eight on left side due to fusion of fifth and sixth), last very large but bordering for only a short distance on mouth; eight infralabials, first pair narrowity in contact behind small triangular mental; four infralabials touching first chinshields which are larger but nearly same length as second pair. Scales smooth without apical pits. 698 Tue UNIVERSITY SCIENCE BULLETIN Scale-formula: 24 (occiput), 17, 17, 17; ventrals, 146, anal di- vided; subcaudals, 32 + (tail defective, part missing). The maxillary teeth small, subequal, 44-44; ptergopalatine, 45- 46; mandibular, 45-40 +, all with brown tips. Color: Body with a median dark blackish-brown median stripe covering three median scalerows and edges of adjoining rows, the stripe darkening anteriorly and posteriorly; on either side of this a light yellow-brown stripe, with a series of dark spots or bars often alternating, becoming darker posteriorly, and nearly black anteriorly; brown below this on sides, the color extending onto ventrals, touch- ing a series of deep-black dots which extends from a point an inch behind head to end of tail. Head and occipital region for a dis- tance of about half an inch (8 scales) blackish with faint light flecks; a black line, irregular in width, on sides of neck; posterior labials cream-white with black edges above and some darker fleck- ing below; anterior labials dark, each with a whitish area; infralabials and entire chin faintly clouded with fine pigmentation; throat and venter yellowish-white. Measurements in mm. (P.K. Nos. 106, 107 respectively): Snout to vent, 288, 128; tail, part missing, 58, 53; width of head, 7.8, 5.5; length of head, 13, 10. Variation: Only two Thai specimens are available to me, the second being a juvenile from the same locality as the one described. There are 148 ventrals; the anal scale is divided, and there are 117 subeaudals. The posterior part of the body shows the darkening as in the described specimen but the head and neck are lighter. The posterior dorsal head scales are light brown, somewhat darker between the eyes, then somewhat lighter on snout. The scales have some whitish spots and their borders are gray or black. The chin is darker than in the described adult. The rows of black spots are indicated but not clearly marked anteriorly. Distribution: In Thailand the species is known only in Narathiwat and Pattani. It is reported as being common in the lowlands of Malaya by Tweedie (Snakes of Malaya, 1954, pp. 63-64). Remarks: Tweedie has stated that Dr. L. D. Brongersma has established the fact that the snake S. geminatus (Boie) is a species distinct from this one and it is probably confined to Java. He states that the correct name for this Malayan snake, which is also found on the islands of the Sunda Shelf, is S. melanocephalus. SERPENTS OF THAILAND 699 Dr. Mertens has described a subspecies geminatus insularis which may actually belong to melanocephalus, (Senckenbergia, vol. 9, 1927, p. 240.). Sibynophis triangularis Taylor and Elbel (Fig. 18) Sibynophis collaris triangularis Taylor and Elbel, Univ. of Kansas, Sci. Bull., vol. 38, pt. 2, Mar. 20, 1958, pp. 1130-1134, fig. 23 (type-locality, Nongko, Sriracha, Chon Buri province, Thailand). Sibynophis collaris (part.) Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 277-278. Diagnosis: Related to Sibynophis collaris; a cream-line from rostral extends along the labials and meets its fellow on median nuchal region, widening posteriorly; postoccipital black spot tri- angular; a median row of quadrangular black spots interspersed with white spots; a lateral light line extends from neck some distance along the fourth lateral scalerow. Description of species (from type description): Rostral three- fifths times as high as wide, narrowly visible above; internasals a little wider than long, their posterior borders curving, laterally narrowed, touching loreal; frontal a little wider than supraocular (4 mm.x2.5 mm.), longer than its distance from end of snout; parietals longer than wide, about equal to their distance from rostral; nasal scale divided or at least partially divided with a vertical depression following nostril, posterior portion higher than anterior, touching two supralabials; loreal small, longer than high, touching two supralabials; a large preocular, higher than wide; ten supralabials, fourth, fifth, and sixth enter orbit; sixth and sev- enth touch lower postocular; a single anterior temporal touches eighth; tenth and last much the largest of series; two postoculars; parietal touching only upper postocular which is nearly square; 1 temporal formula, 1 + ; nine infralabials, fourth largest, 1+1 first four touching first chinshields which are larger than second pair; scales smooth. Scalerows, 17-17-17; ventrals, 160; anal, divided; subcaudals (in- complete), 49. Color: Generally grayish brown with a fawn line beginning on neck at level of eleventh ventral, following fourth and part of fifth scalerows; somewhat farther back it becomes discontinuous form- ing tiny elongate or rounded spots on upper half of fourth scalerow, 700 THE UNIVERSITY SCIENCE BULLETIN Fic. 18. Sibynophis triangularis Taylor and Elbel. Type. KUMNH, No. 33520, Nongko, Sriracha, Chon Buri province, Thai- land. Actual total length, 327 mm. (From Taylor and Elbel. ) SERPENTS OF THAILAND 701 sometimes including part of adjoining scale in fifth row; pigment surrounding these spots becoming somewhat blackish; on middorsal line a row of small black spots extends a short distance; farther back a discontinuous series of white dots begins and continues posteriorly; these finally become nearly obsolete but they can be traced on to tail; ventral surface whitish with a discrete rounded black spot on outer edge of each ventral and subcaudal; outer turned-up part of ventrals pigmented like lateral body scales. Head variegated olive with brownish-black spots, one or two on each head scale; an indefinite transverse dark band across back of frontal and supraoculars. A second somewhat indefinite curving band bordering parietals and temporals, joining deep black tri- angular nuchal spot and this enclosing an olive area; a cream-line from rostral passes back across labials and meets its fellow on medial nuchal region; this bordered above with black on labials; on lower part of some labials black spots present; anterior part of infralabials light brown with darker-edged cream spotting. Maxillary teeth, 49-47; mandibular teeth, 53-52; palatine-pterygoid series, 53-53. Measurements in mm.: Snout to vent, 239; tail (mutilated), 88; head-width, 8.15; head-length, 14. Distribution: Known certainly only from southeastern Thailand in the province of Chon Buri. Malcolm Smith (1943) states: “Specimens [of collaris] from Siam and Annam may have a lateral series of yellow spots on scale-row 4 or 5, and the yellow border on the nape may be chevron-shaped, the apex pointing backwards.” I suspect that these may be referable to triangularis. Remarks: Originally Taylor and Elbel (1958) regarded this a subspecies of Sibynophis collaris. Since the two forms are found together through a wide range it seems wiser to regard each as of specific rank. I have not been able to obtain further specimens of this species. Sibynophis collaris (Gray) (Figs. 19, 20) Psammophis collaris Gray, Ann. Mag. Nat. Hist., ser. 2, vol. 12, 1853, p. 390 (type-locality, Khasi Hills, Assam). Polyodontophis coilaris: Boulenger, Fauna of British India . . . Reptilia and Batrachia, 1890, p. 302; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, p. 184, pl. 12 (part.); Annandale, Rec. Ind. Mus., vol. 8, 1912, p. 46; Wall, Journ. Bombay Nat. Hist. Soc., vol. 18, 1908, p. 316; ibid., vol. 19, 1909, p. 340, 757; ibid., vol. 29, 1923, p. 598; M. Smith, 702 THe UNIVERSITY SCIENCE BULLETIN Journ. Nat. Hist. Soc, Siam, vol. 1, 1914-1915, pp. 129, 155, 212; ibid., vol. 1, no. 4, 1915, p. 244 (Koh Chang); Fraser, Journ. Bombay Nat. Hist. Soc., vol. 39, 1937, p. 498. Ablabes collaris: Stoliczka, Journ. Asiat. Soc. Bengal, vol. 40, 1871, p. 4380. Sibynophis collaris: Smith, Bull. Raffles Mus., no. 3, 1930, p. 40; Rec. Ind. Mus., 42, 1940, p. 482; Pope, Reptiles of China, 1935, p. 86, fig.; Bourret, Les serpents de l’Indo-Chine, 1936, p. 43 (part.); Shaw, Shebbeare and Barker, Journ. Darjeeling Nat. Hist. Soc., vol. 13, 1939, p. 115; M. Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 277-278, fig. 94, A, B,C, D (part.). Diagnosis: One anterior temporal touching eighth labial; all ventral scales with a black dot on outer edges; anterior ventrals also with an inner central pair of dots; nape with a transverse black bar bordered behind with white or yellow, little less than width of a single transverse scalerow; a whitish series of spots on anterior labials becoming elongated dashes, and more or less continuous posteriorly. Description of species (from No. 54, Doi Suthep, Chiang Mai): Rostral twice as wide as high, visible above for half its height; in- ternasals subcircular in form, less than half size of prefrontals which are strongly rounded behind; frontal shield-shaped, longer than its distance from tip of snout, shorter than parietals; supra- oculars longer than frontal; nostri] between two nasals, posterior part highest; loreal a little wider than high; large preocular; eye large, pupil circular, diameter of eye equal to its distance from front edge of nostril; two postoculars; ten supralabials, fourth to sixth bordering eye; eighth and tenth much enlarged; nine infra- labials, mental triangular; first labials in contact; first chinshields larger but about as long as second pair. Scale-formula: occiput (at fifth ventral), 23; neck, 18; middle of body, 17; before vent, 17. Ventrals, 176; anal divided; sub- caudals, 116, divided. Color in life: Head dark olive with a pair of blackish spots sur- rounded by tiny yellow dots on rostral, one each on internasals, prefrontals, and supraoculars; minute yellowish punctations on parts of parietals and frontal (distinct under a lens); dim inter- ocular dark band and a similar band behind occiput, bordered behind by a dotted yellow line, each dot smaller than one scale; ocellated white spots on labials; those on posterior labials elongated blue-white dashes bordered below and above by black, more or less connecting with the narrow yellow line. Body brownish with an indistinct median row of tiny black dots and lighter points be- tween; two median rows a little darker than adjoining scalerows; SERPENTS OF THAILAND 703 edges of ventrals and subcaudals colored like first scalerow with a black dot near outer side of each; anterior ventrals with a median pair of black spots also. Measurements in mm.: Total length, 712; snout to vent, 530; tail 182; width of head, 8.7; length of head, 15.5. Variation: The coloration in young animals is more distinct, and the pattern more sharply delineated than in old ones. A figure of a young specimen, No. 3180 (Chiang Mai province) is given. It brings out the character of the striping in the posterior part of the body which is largely lost in old adult specimens. The head markings are much the same in these two, as well as in No. 36070 captured at Kaeng Pang Tao in northern Chiang Mai (figured). Fic. 19. Sibynophis collaris (Gray). No. 3180, Young, Chiang Mai province, Oliver Gordon Young, Coll. Actual length (tail mutilated), snout to vent, 310 mm. THe UNIVERSITY SCIENCE BULLETIN Fic. 20. Sibynophis collaris (Gray). No. 36070, Kaeng Pang Tao, northern Chiang Mai. Actual total length, 565 mm. SERPENTS OF THAILAND 705 The known variation in ventrals (fide M. Smith) is 155 to 186. Distribution: The specimens of this species that I have found or been able to study are all from Chiang Mai province. Dr. Malcolm Smith report the species from Koh Chang (island) in the Gulf of Siam. Elsewhere collaris has been reported in northern India, Assam, Nepal, northern Burma and China. Subfamily ACROCHORDINAE Jan Acrochordidae Jan, Elenco syst. ofid., 1863, p. 106. Acrochordinae Boulenger, Catalogue of the snakes in the British Museum, ed. 2, vol. 1, 1893, p. 172 (part.). Hypapophyses present throughout vertebral column; postorbital bone semicircular, partly surrounding upper part of orbit; frontal with two anterior lateral processes; prefrontal small, vertically hang- ing from the end of frontal expansion, not extending forward upon snout. Skin loose with or without median ventral “seam” or fold. While there is considerable reason for recognizing these snakes as a distinct family, they are here treated as a subfamily. For many years two genera, Chersydrus and Acrochordus were recognized in this subfamily, the most significant difference be- tween them being the presence of the median ventral “seam” or fold present in Chersydrus. M. Smith (1943) has united the two stating that, “The presence of a distinct median abdominal fold in Chersydrus granulatus does not seem sufficient to separate it gen- erically from Acrochordus.” There are, however, certain other differences besides this (e. g. ear-bones). However, until further study of the skeletal char- acteristics demonstrate still greater differences, I shall accept the proposal of Smith and regard Acrochordus as the only genus in the subfamily, with Chersydrus a synonym. Genus Acrochordus Hornstedt Acrochordus Hornstedt, Vet. Acad. Hand., Stockholm, vol. 8, 1787, p. 307 (type of genus, javanicus). Potomophis Schmidt, Abh. Naturv. Hamburg, vol. 2, 1852, p. 75 (type javanicus; name preoccupied ), Chersydrus Cuvier, Régne Animal, vol. 2, 1817, p. 75 (type, fasciatus Shaw = granulatus Schneider). Diagnosis: Head not distinct from neck; body moderate to very stout, the skin loose, covered with small juxtaposed tubercles; no ventrals present; nostrils surrounded by circular scales; eyes dorsal, 28—1367 706 THE UNIVERSITY SCIENCE BULLETIN small; pupil vertical; chin produced forward fitting into a depres- sion in upper jaw. Maxillary teeth, 12-15, subequal; tail only feebly compressed, somewhat prehensile. The distribution of this subfamily extends from India through southeastern Asia, the Indo-Australian Archipelago into Northern Australia. Two species are recognized, both occurring in Thailand. They are found in marine waters, river mouths, and even in fresh water. In Thailand the species javanicus is known as ngu nguang chang, the elephant-trunk snake. It is not uncommon in the canals of Bangkok. KrEy TO THE SPECIES OF ACROCHORDUS IN THAILAND 1. Nostrils terminal directed forward; from 130-150 scales about body, no distinct ventral median fold; no transverse bands; usually some longi- tudinalistripimg ii. jo: sje wre ee ila ook eee Oe eee javanicus Nostrils dorsal, directed upward; approximately 100 scales about middle of body. A well-defined median ventral fold; a banded color pattern, grcnulatus Acrochordus javanicus Hornstedt Acrochordus javanicus Hornstedt, Abh. Acad., Stockholm, vol. 8, 1787, p. 307 (type-locality, Java); Schlegel, Abbildungen . . . Amphibia, 1839, pl. 17, figs. 12-14; Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 905; Giinther, The reptiles of British India, 1864, p. 336; Boulenger, Catalogue of the snakes in the British Museum, 2nd ed., vol. 1, 1893, p. 173; Flower, Proc. Zool. Soc. London, 1899, p. 658; Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, 1914-1915, p. 13, 187, 212; de Rooij, The snakes of the Indo- Australian Archipelago, vol. 2, Ophidia, 1917, p. 48, fig. 22; Smith, Bull. Raffies Mus., no. 3, 1930, p. 39; Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 132-134, fig. 43, a and b; Tweedie, The snakes of Malaya, 1954, p. 38. Diagnosis: Large snakes reaching approximately two and one- half meters in length, the girth reaching above 300mm.; snout blunt, nostrils directed forwards and upwards; eyes dorsal; tuber- cular scales small, 130-150 around middle of body; no ventral fold on body; no widened ventrals. Description of species (from No. 1693, Bangkok): Front of snout covered with several curving rows of small irregular scales; nostrils surrounded by two ringlike scales touching mesially, directed straight forward; eyes small, separated from one another by 23 scalerows, from nasal by 14 scalerows; about 30 supralabials, sep- arated from eye by ten scalerows; series adjoining supralabials a little larger than labials; about 31 infralabials; skin of lip itself thick and broken into more or less regular scalelike parts; median part of lower jaws forms a small forward projection that fits into a small arch in upper jaw. Scales on head smallest above eyes and in mid- dle of occiput. SERPENTS OF THAILAND 707 Posteriorly on tail scales rather flattened with a median keel and each side elevated into two blunt keels. These scales may be imbricate; anteriorly keels become higher and somewhat mucronate, the scales separated, tending to stand more erect; posterior edge of scale develops some small spinelike projections; finally scale stands erect the median keel now like a small cone with lateral mucrone- like projections forming smaller cones. The central part around cones covered with perhaps hundreds of fine hairlike projections visible under magnification. On median ventral line a vague elevation; tubercles here smaller than elsewhere around body, those on chin and throat smallest of all. Scales around neck, 126; about middle of body, 126; just anterior to vent, 58. Approximately 566 transverse rows; the ex- truded hemipenes are approximately 110mm. in length. Each bifurcates 38 mm. from end, the entire bifurcated parts covered with about 15 somewhat irregular rows of short hard spines. The groove divides and one branch reaches terminus of each bifurca- tion; proximal to the division there are smooth irregular folds. Color in life: Above generally olive with a black median stripe with irregular borders; below the olive color a pale yellow stripe with poorly defined edges; below this is a black stripe broken at intervals, continuing back onto tail; below black stripe the sides and venter dull yellowish white, and about midway of body two or three series of dark spots of irregular shape and size begin and con- tinue to vent; subcaudal region dull olive-brown without spots. Measurements in mm.: Total length, 1725; tail, 301; snout to vent, 1424; width of head, 36; length of head, 55. Variation: The number of supra- and infralabials vary between 25-32; the number of scales between the eyes varies between 18 and 24; and the number of scales around the body, between 130 and 150. The young are more clearly marked than the adults, and they may be spotted all over. Distribution: Known from mouths of streams emptying into the Gulf of Siam, and in canals and fresh water klongs that are near stream mouths. It is not a rare snake in klongs and canals in Bangkok. It is known from the coasts of Cambodia, Cochin China, Malay, the Indo-Australian Archipelago, and Queensland. Remarks: These snakes are known to reach a length of 2304 mm. (fide Giinther). They produce 25 to 32 living young at a time. The food consists largely or entirely of fish. 708 THE UNIVERSITY SCIENCE BULLETIN Acrochordus granulatus (Schneider ) (Fig. 21) Hydrus granulatus Schneider, Historiae amphibiorum naturalis et literariae, vol. 1, 1799, p. 243 (type-locality, India). Chersydrus_ granulatus: Gimther, The fauna of British India, 1864, p. 336; Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, Saigon, 1885, p. 39; Boulenger, The fauna of British India, Reptilia and Batrachia, 1890, p. 355, fig.; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, p. 174; Flower, Proc. Zool. Soc. London, 1899, p. 658; Boulenger, A vertebrate fauna of the Malay Peninsula ite Reptilia and Batrachia, 1912, pp. 116-117, fig. 38: Annandale, Journ. Asiat. Soc. Bengal, vol. 74, 1905, p. 175: Mem. Ind. Mus., vol. 5, 1915, p. 169; Wall, Journ. Bombay Nat. Hist. Soc., vol. 15, 1918, p. 756; The snakes of Ceylon, 1921, p. 79; Prater, Journ. Bombay Nat. Hist. Soc., vol. 30, 1924, p. 167; Brongersma, Zool. Med. Rijks. Mus., vol. 16, 1933, p. 14; Cochran, Proc. U.S. Nat. Mus., vol. 77, 1930, p. 14: Bourret, Les serpents de l’Indochine, 1936, pp. 35-36, fig. 14. Acrochordus fasciatus Shaw, General zoology, vol. 3, p. 576, pl. 130 (type- locality not stated). Chersydrus annulatus Gray, Catalogue of the snakes in the collection of the British Museum, 1849, p. 61 (type-locality, Singapore and Madras). Acrochordus granulatus: Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 906; M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3 Serpentes, 1943, pp. 134-135; Tweedie, The snakes of Malaya, 1954, p. 38; Smith, Bull. Raffles Museum, no. 3, 1930, p. 39. Diagnosis: Body whitish with a series of 50 or more black bands, widened on back, narrowed on sides and venter; no widened ven- trals; a narrow median fold on venter; eye very small with a minute round pupil; nostrils on top of snout, directed upwards; tail some- what compressed; large series of scales on upper and lower jaw separated from mouth by a series of much smaller scales. Scales tubercular or keeled; about 100 scales around middle of body. Description of species (from No. 479, Songkhla): Rostral scale very narrow, doubly arched, concealed in a notch of upper jaw, not visible from above; twelve small scales border upper lip; bordering these scales above, a series of 12-13 larger (true) labials; between nostrils and the enlarged series there are two rows of scales, the lower ones enlarged; a single row of scales between nasals; nasals are narrow circular rims about circular craters, followed behind by at least one enlarged scale; eye small with a minute circular pupil, surrounded by eleven small scales varying in size; approximately twelve scales in a line between eyes; a series of about twelve small scales separating a series of much larger in- fralabial scales from lip; mental in notch between lower jaws, median part of lower jaw fitting into a median notch in upper jaw; scales on snout behind nostrils larger than those on occiput. Fic. 21. Acrochordus granulatus (Schneider). No. 479, Songkhla, Songkhla province, southern Thailand. Actual total length, 649 mn tail, 69 mm. 710 THe UNIVERSITY SCIENCE BULLETIN Scalerows about body: 62, 77, 93, 83, 48, largest scales on dorsum, each with a keel or tubercle; scales along median part of venter pointed; most scales at least anteriorly on body, are elevated, with a short spine arising near to a terminal pit. Maxillary teeth, 16, for the most part subequal, smallest pos- teriorly; 15 mandibular teeth decreasing posteriorly. Color in life: Head blackish with a few dirty white spots; neck and body dirty whitish encircled by approximately 52 black bands, widest on middorsal line, narrowing on the sides and _ venter, often anastomosing on back, tending to form a median zigzag stripe for a distance. Tail with 13 bands. Measurements in mm.: Total length, 649; tail, 69; head-width, 14; head-length, 16.5. Variation: The markings become indistinct in old specimens. The scales in a row between the eyes vary from seven to ten. The scales around the middle of body vary between 89 and 100. Distribution: In Thailand the species is not uncommon along the coasts. Specimens are more frequently taken near mouths of rivers emptying into the Gulf of Siam. Outside of Thailand the species is known in India, Ceylon, Indo-China, Malaya and the Indo- Australian Archipelago as far as Australia and the Philippines. In the latter country the species occurs very commonly in the large fresh-water lake Laguna de Bay, on the Island of Luzon. Remarks: This serpent seemingly pays no attention to the pres- ence of man. They are quiet and inoffensive. I have never seen one attempt to bite. Subfamily CoLuprinaE Cope Key to THat GENERA OF THE COLUBRINAE iw) 1., Panglike anterior maxillary, teeth... -. 15.0 = ee eee No fanglike anterior maxillary teeth......................0220-000- 3 2. Maxillary bone strongly arched; rather small snakes (usually under one TMS FEET eT Ot bays eh espn Cesta deere Nec ae ee ere ay eco a Lycodon Maxillary bone not strongly arched; larger, (to one and one-half meters, Dinodon 3. Maxillary teeth increasing in size from front to back.................. 4 Maxillary teeth subequal or anterior teeth slightly larger............. 6 4, Pupil vertically elliptic; six to ten teeth on maxillary; scales in 13 rows around body; slender banded arboreal snakes............ Dryocalamus Pupilsround 2532 ee oe ead OES, a 5 5. Maxillary teeth 20 to 28 (Thai species), forming a continuous elles, tyas Maxillary teeth 6 to 16, the posterior ones strongly enlarged and com- pressed; rostral large, often elevated extending well on upper surface of PHS ISMO Ube eas ee REN oe ee Eee eee ea eee Oligodon 10. 11. SERPENTS OF THAILAND qeNiontemporalscalles sys ce rere ieee nek ees ees eres ee eeneeh (emporalescalessoresemtign 5) creda cre Gee tos ae cae aa, . Ten to twelve maxillary teeth; very small snakes (250 mm.); internasals present; scales in 15 rows..........:........0.0.0005. Pseudorhabdion Eight to eleven maxillary teeth, small snakes, less than half a meter in length; no separate internasals (fused with prefrontals) ; scales in 13 rows, Calamaria Me Scalesmnial OForsmmOnresl OWSEe eis eee eee ee Scaleseiaul 3s orl pro west eet eet ees nee get Meena nantes kee ie ed mene Scalerows, 23-25; snout projecting; anterior maxillary teeth somewhat larger than others; an elongate loreal three to four times as long as high; larcex(2imire ters) ewes feck geht eae e Bae elias, eae Bae ae Oli a Gonyosoma Sealerows, 19-23; maxillary teeth subequal, sometimes anterior, some- times posterior a little the larger; loreal not twice as long as high; ter- restrial snakes, 114 to 2 meters; longitudinal or banded markings; Elaphe Elongate slender arboreal species; scales variable in shape; transverse rows strongly oblique; pupil round; about one meter in length; Ahaetulla Shorter terrestrial species, the scales nearly uniform; pupil round or Verbicallyellip ticker ces ter eee ee ae eee OR een ere She . Pupil vertically elliptic; 16-20 equal maxillary teeth; anterior temporal not longer than posterior; loreal not especially small; (less than half a Meterimeleng thy) et paren een aes Sree ie Ohe iee eae are Plagiopholis Pupil round; 26-28 maxillary teeth, subequal; anterior temporal twice length of posterior; loreal very small; (to half a meter in length); Liopeltis Genus Gonyosoma Wagler 711 10 11 12 Gonyosoma Wagler, Natiirliches System der Amphibien, 1828, pl. 9 (type of genus, viridis = oxycephalum); Taylor, Snakes of the Philippine Islands, 1922, p. 152. Diagnosis: A genus of large arboreal snakes reaching a length of 2.3 meters; scales faintly keeled or smooth; elongate apical pits pres- ent; loreal elongate; hemipenes with spines elongated. There is no unanimity of opinion as to the number of species in this genus, but two forms, possibly three are known to occur in Thailand. It is probable that floweri and jansenii should be placed in the genus Elaphe. 1. Key To ForMs OF Gonyosoma IN THAILAND Uniform green on body, (the interstitial skin with black and white retic- ulations); ventrals outside lateral keel whitish; tail often colored differ- ently from body, usually reddish or each caudal scale edged with darker; tworsupralabials) borderorbitiens see ee ee oxycephalum INotaumitormeereeniye shen eu sia een he ee Tule oe ea at . Light or dark buff marked all over by scattered irregular blotches of dark brown or black; venter uniform whitish or with spots similar to those on dorsum; two supralabials enter eye, third touches it at a point... ... floweri Bright yellow-brown above, the scales black edged. A pair of black lat- eral stripes begin at about the 36th ventral level, and continue back, widening greatly. Venter yellowish anteriorly, blackish with yellow keels posteriorly; three labials border orbit............. jansenii elegans Ww FAQ THE UNIVERSITY SCIENCE BULLETIN Floweri is known from Singapore north to Trang province in Thailand. Jansenii elegans is known from the type taken reputedly in Thailand by (Schilling). Jansenii is a Celebean species, and the presence of this form in Thailand should be reconfirmed. Gonyosoma floweri (Werner ) Coluber floweri Werner, Stizb. Akad. Wiss. Wien, Abt. 1, vol. 134, p. 55 (type- locality, Singapore ). i Elaphe oxycephala (part.) M. Smith, The fauna of British India, Ceylon and ae . . . Reptilia and Amphibia, vol. 3, Serpentes 1948, pp. 144- Diagnosis: See key. Description of species (from type-description): Internasals three fifths as long as prefrontals; frontal one and one-third times as long as broad, as long as its distance from rostral, shorter than parietals; loreal three times as long as broad (left one split vertically); a large preocular in contact with frontal; two postoculars, upper dis- tinctly larger than lower. Temporals, (left), 2+ 2+ 3; (right), 3+2+3-+2-+ 1; ten supralabials, sixth to eighth border orbit; five to seven infralabials in contact with anterior chinshields. Eye not half as long as its distance from nostril. Color: Above bright yellow-brown, scales black-edged with small irregular black flecks. A dark postocular band above upper edge of supralabials; venter yellowish; tail spotted. Measurements: About one and one-half to two meters. Remarks: This snake is related to Gonyosoma j. jansenii Bleeker of Celebes. The three supralabials, sixth to eighth (fifth to seventh) borders orbit. The postorbital band is present and this and the distinctive coloration of the posterior part of the body seems to separate floweri from the species in Celebes. Gonyosoma jansenii Bleeker Gonyosoma jansenii Bleeker, Nat. Tijdschr Nederland Ind., vol. 16, 1858; (type- locality, Manado, Celebes); Boulenger, Proc. Zool. Soc. London, 1897, p. 220. Allophis nigricaudus Peters, Monatsb. Akad. Wiss. Berlin, 1872, p. 686. Elaphis nigricaudus: Peters and Doria, Ann. Mus. Civ. Genova, vol. 13, 1878, p. 387. Diagnosis: See key. This form presumably has two subspecies the typical form in Celebes and a second one described from Siam by Werner as Coluber jansenii elegans. He however regarded it as a “Firben- SERPENTS OF THAILAND li varietit.” There may be some doubt that the form actualiy origi- nated in Thailand. (The collector's name is given as “Shilling”. ) Gonyosoma jansenii elegans Werner Coluber janseni elegans Wemer, Sitzungsb. Akad. Wiss. Wien, Abt. 1, vol. 134, 1925, pp. 55-56 (type-locality, “Siam” ). Elaphe oxycephala M. Smith, The fauna of British India, Ceylon and Burma vol. 3, Serpentes, 1948, pp. 144-146 (part.). Description (from type-description of g ): Ventrals, 257, sub- caudals, 132 + 1; scalerows, 25. Loreal twice as long as high; nine supralabials, fifth to seventh border eye, ninth very long. Temporals, 2+3+4. Color: Bright yellow-brown; lateral stripes begin on sides above the 3rd ventral; at first two small short lines that fuse and widen, covering about eight scalerows posteriorly. The edges of stripe ir- regular at first then they become parallel, the five scalerows be- tween being lighter; posterior part of body and tail entirely black. Anterior part of venter uniform yellow, posteriorly bordered with black. Measurements: Not recorded (“groses” ¢ ). Gonyosoma oxycephalum (Boie in Boie ) Coluber oxycephalus Boie in Boie, Isis, 1827, p. 537 (type-locality, Java); Boulenger, The fauna of British India, ‘including Ceylon and Burma; . Reptilia and Batrachia, JEkO. p. 335; Catalogue of the snakes in the British Museum, 2nd ed., vol. 2, 1894, p 56: Annandale, Journ. Asiat. Soc. Bengal, vol. 1, 1905, p. 75: Wall and Tae Journ. Bombay Nat. Hist. Soc., vol. Se 1901; Boulenger, A vertebrate fauna of the Malay Peninsula , Reptilia and Batrachia, 1912, pp. 143-144; Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 622; M. Smith, Proc. Zool. Soc. Legato 1921, p. 426: ’ Rendahl, Ark.-Zool. Sven. Vet., vol. 29, A. 10, 1987, p IAAI ES oxycephalus: Schlegel, Essai sur la a Gone We serpents: vol. 1837, p. 189, pl. VII, figs. 8, 9; Cantor, Journ. Asiat. Soc. Bengal, vol. 4 1847, p. 927, Gonyosoma oxycephalum: Giinther, The reptiles of British India, 1864, p. 294. Stoliczka, Journ. Asiat. Soc. Bengal, vol. 39, 1870, p. 193. Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, 1885, pp. 51, 65; Taylor, Snakes of the Philippine Islands, Bureau of Sci. Manila Publ. 16, 1922, pp. 152-155; Schmidt, Amer. Mus. Novitates, no. 157, 1925, pp. 1-5; Dowling, Copeia, 1958, no. 1, pp. 29-40, fig. 1A. Elaphe oxycephala: M. Sma, Journ. Nat Hist. Soc. Siam, vol. 1, no. 3, Mar. 1915, p. 218; ibid., vol. 2, No. Dec. 1916, p. 161; Proc. Zool. Soc. London, 1921, p. 426: Journ. Peace Malay States, ‘vol. 10, 922s: 266. Bull. Raffles Mus., no. 3, 1930, p. 50; Bourret, Les serpents de l’Indo- chine, 1936, vol. 2, pp. 9204-206, fig, 81; M. Smith, The fauna of British India, Ceylon, and Burma... Reptilia and Amphibia, vol. 3, Ser- pentes, Dec. 1943, pp. 144-146, fig. 45. Gonyosoma viride Wagler, Descriptiones et icones amphibiorum pt. 1, 1828, pl. 9 (type-locality “Brazil” ). Alopecophis chalybeus Gray, Ann. Mag. Nat. Hist., ser. 2, vol. 4, p. 247 (type- locality “Mauritius” ). 714 THE UNIVERSITY SCIENCE BULLETIN Aepidea robusta Hallowell, Proc. Acad, Nat. Sci. Philadelphia, 1860, p. 488 (type-locality, Gasper Straits, Malay Archipelago); Stejneger, Proc. United States Nat. Mus., vol. 69, 1926, p. 3 (= oxycephalum). Diagnosis: Large snake exceeding two meters in length; snout about three times as long as eye; supralabials, 10 to 11, last much enlarged; 12 to 14 infralabials; ventrals somewhat keeled; scale- formula: 23, 23, 15; entire head and body green in life; tail brown- buff or reddish; light yellow-green below. Description of species (from No. 33638, Doi Suthep, Chiang Mai, Gordon Young, Coll.): Snout three times length of eye, some- what wedge-shaped extending noticeable beyond mouth; rostral higher than wide; visible above as a small triangle; internasal longer than wide; prefrontals longer than wide, touching loreal and posterior nasal laterally; frontal small, rounded posteriorly front edge nearly straight anteriorly, less than a fifth longer than wide, shorter than its distance to end of snout; shorter than parietals; nasal completely or partially divided, anterior part as large as posterior; loreal about three times as long as high; preocular large, about as long as high, bordering frontal; supraoculars three fifths as wide as frontal; two postoculars; temporals somewhat irregular; nine supralabials, three touching loreal, sixth and seventh entering orbit; last double the length of two preceding it; infralabials, 12, six touching large chinshields which are more than four times size of second pair. Scale-formula: 30, at first ventral, 23 (24), 24 (25), 16, (15). The scales with indistinct paired apical pits; several dorsal scale- rows very near middle of body showing almost obsolete keels. Each scale shows longitudinal striation under magnification. Maxillary teeth 21-21; anterior a little longer and more curved than posterior. Color in life: Body green above, darkest on head; tail light chest- nut or buff-red, the two colors meeting abruptly at level of vent. On anterior half of body scales may be edged with black; an in- distinct blackish stripe along side of head immediately above labials; light greenish-yellow below on venter. The specimen here described is darkened by preservation and is nearly uniform plumbeous; venter dark, the posterior edge of each ventral scale being whitish; a whitish line following blunt keel along each side of ventrals. A light area on rostral and a light area on each side under head. The chinshields and infralabials plum- beous; a slightly wider light Ine follows keel on subcaudals; tail brownish, the scales with a somewhat lighter center. SERPENTS OF THAILAND aS Measurements in mm.: Snout-vent, 447; tail, 142; total length, 589; length of head, 20; width of head, 12. Variation: A specimen captured at Khao Chong (No. 35708) dif- fers somewhat from that described. The rostral is visible above as a narrow line; the internasal is small, less than half the area of a prefrontal. The frontal is pentagonal and rounded somewhat pos- teriorly. Two (or three) preoculars, the one (or two) lower scales separate the fifth labial from eye. The temporal formula is 1+2-+3. Ten supralabials are present, the sixth and seventh bordering eye, and thirteen infralabials. The second pair of chin- shields are separated by two or three scales. The coloration in life of this specimen was as follows: The body was green with a buff or reddish tail, and a greenish yellow venter. Many of the scales on the anterior part of the body were decorated with a tiny line of blackish dashes, while farther back many of the diagonal transverse scalerows had each scale bear- ing tiny white dashes. On the sides the black and white marks were concealed normally by imbricating scales. The top of the head and chin were green while the throat was yellowish white. The tail was reddish but each scale had a darker border. A median dark zig-zag line flanked by unpigmented reddish stripes, was on under side of the tail. This specimen has the following scale-counts: scale-formula 32 (occiput), 27, 25, 25, 17. The ventrals are 249, keeled and often notched; subcaudals, 138; anal, divided. Distribution in Thailand: The species has been taken in the provinces of Chiang Mai, Narathiwat, and Trang. Outside of Thailand the species is known in Cambodia, Malaya, Singapore, Burma, Borneo, Java, and the Philippines. Remarks: The species is very large reaching a known length of 1880 for a male; 2100 mm. for a female. Genus Elaphe Fitzinger in Wagler Elaphe Fitzinger, in Wagler, Descriptiones et icones amphibiorum, pt. 3, 1833, text to pl. p. 27 (type of genus parreysi = quatuorlineatus ). Diagnosis: Maxillary teeth, 14-24 somewhat enlarged anteriorly and posteriorly; head only slightly distinct from neck; pupil round; body subcylindrical; scales with paired apical pits in 19 to 27 rows, smooth or keeled; ventrals rounded or angulate laterally; sub- caudals divided; five or six infralabials touch first chinshields. 716 THe UNIVERSITY SCIENCE BULLETIN A very large genus with species in Asia, Europe, and North America. In Thailand four species are known, all of them having wide ranges. They may be differentiated by the following key: Key To THAI Species OF Elaphe 1 3Scales\in 19)\G@nore! rarely 21) ows. 4s sd) es ee 3 Scales:im more thant’ 21 -rowse: fission eeeee 2. Seales in 23 rows; a black stripe along side of head; anterior part of body nearly unicolor; posteriorly with a median light stripe and two broad DVaACK StI POS a site icone aia aera ect Premera eee aaa taenivura ridleyr Seales in 23 to 25 rows; anterior part of body with a median series of butter- fly-shaped spots and a lateral series of smaller diamond-shaped spots, taeniura taeniura 3. A single narrow black transverse occipital bar or band with black lines running forward from bar to eye; head coppery brown.......... radiata Numerous wide dark transverse bars on body or none................. + 4. Seales smooth; body reddish brown with numerous wide dark-edged dark- brown crossbands; a pair of black lines running forward from first band to eyes; a median black stripe on head........... porphyracea porphyracea Seales keeled; a median light yellowish-brown stripe flanked by two nar- row black stripes anteriorly; a series of black ventrolateral spots or ocelli on: anterior third of body.) ee ee ae ee flavolineata bo Elaphe taeniura (Cope) Elaphis taeniurus Cope, Proc. Acad. Nat. Sci., Philadelphia, 1861, vol. 12, p. 565 (type-locality, Ningpo and Siam); M. Smith, The fauna of British India . . . vol. 3, Serpentes, 1943, pp. 150-152. Thailand has two forms, presumed subspecies of taeniura. One, t. ridleyi, is a common inhabitant of certain caves in peninsular Thailand; the other a northern form, reported by Malcolm Smith, from Chiang Mai and Hin Lap in the Dong Rek mountains, pre- sumably belongs to the typical subspecies. Key To THat Susspecies OF Elaphe taeniura 1. Top of head dark gray; side of head with a black stripe from angle of jaw, fading near nasal. Anterior part of body pale flesh gradually developing pigment forming two broad black bands with a median cream band be- tween and bordered by yellow stripes on outer scalerow and outer part of ventrals; this yellow line bordered by faint dark lines to tip of tail, t. ridley . Similar head marks but anterior part of back with a vertebral series of butterfly-shaped black spots and smaller diamond-shaped ones on sides; posterior part of body similar........ Gin lia Seaai Neeiee Sigh ee eae ..t. taentura iw) Elaphe taeniura ridleyi ( Butler ) (Fig. 22) Coluber taeniurus var. ridleyi Butler, Journ. Bombay, Nat. Hist. vol. 12, 1899, p. 426 (type-locality, Batu Caves, Kuala Lumpur, Malaya); Ridley, The Times, Nov. 10, 1937. Coluber taeniurus: Boulenger, Fasciculi Malayenses, vol. 1, zool. 1, 1903, pp. 162-163 (“Goah Tanah, near Biserat, Jalor’” ). SERPENTS OF THAILAND He: Coluber taeniurus pallidus: Rendahl, Ark. Zool. Stockholm, vol. 29, (A), 10, 1937, p. 19 (type-locality, Sukli, Tenasserim, Burma). Elaphe taeniurus grabowskyi: (part.) M. Smith, Bull. Raffles Mus., Sinagpore, Straits Settlements, no. 3, Apr. 1930, pp. 49-50. Elaphe taeniura: Smith, (part.), The fauna of British India, Ceylon and Burma, including the whole of the Indo-Chinese subregion; Reptilia and Amphibia, vol. 3, Serpentes. Dec. 1943, pp. 150-152. Diagnosis: Anterior maxillary teeth larger than posterior. Head dull slate above with a black stripe through loreal region to jaw- angle; supralabials and infralabials generally yellowish. Anterior half of body light fawn, the neck region only a shade darker; near middle of body, dorsum and sides, except outer scalerow and adjoining half-row, with more dark pigment, continuing to grow from grayish slate to black except for a broad yellowish median line. Description of subspecies (from No. 665 from Khao Kampan, Yala province, Thailand): Rostral wider than high not or scarcely visible from above; internasals tending to bend down slightly an- teriorly, equal in size to about one half area of prefrontals which have a common suture nearly the length of the scales; frontal (9 mm.x1ll mm.) shorter than its distance from tip of snout, shorter than its distance to end of parietal; two nasals, subequal in size; loreal elongate, one-and-a-half times as long as high; two preoculars, lower small (subocular), upper large forming a suture with frontal; supraoculars subtriangular; parietals bordered by three temporals; two postoculars; temporals, two anterior, arranged LD 1 sixth bordering orbit; infralabials, 11-12, six touching first chin- shields which are a little wider but about same length as second pair. Scale-formula: 32 (occiput), 29, 27, 19. Scales smooth on anterior part of body, becoming weakly then strongly keeled except on two or three outer rows which remain smooth throughout. Scales with paired apical pits. Body compressed posteriorly; venter with lateral keels; outer edge of ventrals turned upwards on side; keels not in evidence on neck and tail. Ventrals, 295; subcaudals, 114; anals, 2. Color in life: Head dull slate above with a black stripe through loreal region and back of eye to jaw-angle; supralabials yellow ex- cept for first two or three, and upper edges of certain others; in- fralabials entirely yellowish; anterior half of dorsum yellowish fawn except outer part of ventrals; near middle of body pigment becomes evident, at first light lavender, then becoming darker lavender, + 3 (4); supralabials, 9-9, three touching loreal, fifth and 718 THe UNIVERSITY SCIENCE BULLETIN Fic. 22. Elaphe taeniura ridleyi (Butler). EHT-HMS No. 1627, Khao Kampan (Mt.) cave, Yala province, southern Thailand. Actual total length, 1852 mm.; tail, 370 mm. purplish slate, and finally black towards end of tail; about midway, scalerow bordering lateral part of ventrals and half of adjoining scalerow develops a bright yellow stripe which includes outer edges of ventrals; below this a dorsoventral dark line on upturned part of ventrals continues to end of tail on outer part of subcaudals; a median yellow stripe begins on dorsum at about the middle of body and grows more intense, covering one whole and two half rows of scales, extending to tip of tail; venter yellowish; underside of chin ivory. Measurements in mm. (No. 665 and an unnumbered specimen in the Boonsong Lekagul collection, both from same cave): Snout to vent, 1475, 1492; tail, 365, 390; total length, 1840, 1882; diameter of eye, 6, 7; eye to end of snout, 15, 15; width of head, 26.5, 32.6; length of head, 46.5, 50. SERPENTS OF THAILAND 719 Variation: The second specimen has 286% ventrals, 114 sub- caudals; divided anal; supralabials, 9-9; infralabials, 11-12, five or six touching first chinshields. The color is nearly identical with that of the described specimen. Other head-scales are practically the same, except the temporals have a slightly different arrange- ment. The variation in ventral and subcaudal counts is from 276 to 305; 89 to 112. The scalerows about the middle of the body are 25. While three supralabials normally border the orbit, the fourth may be separated only by a small subocular scale. This form is usually very pale. Those found in the open are somewhat darker but presumably the normal habitat for this sub- species is in caves. All cave specimens I have seen are extremely uniform in color and markings. Distribution: The pale ridleyi is known in Thailand in the south- ern part of the peninsular area. Specimens have been taken only from caves in Yala province. The subspecies is also known from the limestone caves of north- ern Malaya, and it may be presumed that Rendahl’s taeniura pallidus from Sukli, Tenasserim belongs here. The Bornean form of taeniura has different anterior coloration and markings; also fewer ventrals and subcaudals than the penin- sular form. This subspecies has been named ft. grabowskyi. Remarks: The specimen described is from a cave in Khao Kampan in this district of Yala. The first record is that of Boulenger from “Goali Tanali, near Biserat, Jalor.”. This place name seemingly is unknown to the area at the present time. The entire political geography has changed. The seven Pattani States have been divided into the present Changwats of Narathiwat, Pattani, Yala and Songkhla and it is not possible to state with certainty whether Jalor is wholly within the present province of Pattani or not. I visited the Khao Kampan cave with Nai Prayoon Kananuracks. We obtained three specimens of this species. While here we came upon a Dendrasphis hannah (the king cobra) at a distance of more than 350 meters back from the mouth of the cave. I captured the specimen alive with some difficulty. It was presumed that the cobra was seeking the elaphid snakes for food. Elaphe taeniura taeniura (Cope) Elaphis taeniurus Cope, Proc. Acad. Nat. Sci. Philadelphia, 1861, p. 565 (type- locality, Ningpo and Siam). 720 THE UNIVERSITY SCIENCE BULLETIN Coluber taeniurus: Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 333; Catalogue of the snakes in the British Museum, vol. 2, 1894, p. 47. Elaphe taeniura: Pope, The Reptiles of China, 1935, pp. 271-277, fig. 58; M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3; Serpentes, 1943 pp. 150-152 (part.). Diagnosis: Anterior maxillary teeth largest; light brown above uniform on head and neck, except for a black stripe on each side of head, broadest behind eye; anterior part of back with a vertebral series of large black butterfly-shaped spots and smaller diamond- shaped spots on sides which tend to break up with age; posterior part of back with a gray vertebral stripe bordered by a broad black stripe on each side. Description of species (type description): “Head slightly dis- tinct, lanceolate, muzzle obtuse. Tail one-fifth the total length, flat beneath. Twenty-five rows of scales, those from the ninth to the sixteenth keeled. Rostra] plate broader than high, the labial suture one-third the nasal, less than the prefrontal. Eight superior labials, fourth and fifth bounding the orbit. Seventh much longer than high, bounded above by a long temporal, and by a short one, which also bounds the eighth labial. Postoculars two, superior largest. Preoculars two, as in other species of the genus, the superior very large, its horizontal diameter greater than the length of the loreal. The latter plate much longer than high. Postfrontals large, bent upon the sides. Anterior border of the vertical [frontal] shorter than the straight convergent lateral; posterior angle obtuse. Super- ciliaries large. Occipitals [parietals] elongate, external borders convergent, bounded by two long temporals. Inferior labials ten, eleven, or twelve. Gastrosteges [ventrals] 232; one divided anal; urosteges [subcaudals] 101 pairs. Total length of specimen from Ningpo 64 in., tail 13 in.; specimen from Siam 30 in., tail 6 in., 3 lines.” Color: “Above, an olivaceous ash, or clay color, more olive anteriorly. A blackish lateral band extends from the tip of the tail, throughout the posterior third of the body, where it extends from the second to the ninth rows of scales, reducing the ground color to a dorsal stripe of three or four scales in width. It is divided by a number of irregular narrow vertical lines, at regular distances. The superior border is prolonged upon the anterior two-thirds of the body as an irregular, narrow, longitudinal black band, connected with that of the opposite side by similar short transverse bands at SERPENTS OF THAILAND TPAl distances of four or five scales. Irregular black borders and centers of the median lateral scales, are the only indications of the inferior part of the lateral band anteriorly. Gastrusteges tipped with black anteriorly; the central parts become gradually darker posteriorly, but finally give place to a yellow median band which extends to the tip of the tail. This is bounded by a blackish band on each side, which is separated from that of the sides by another yellow one, which involves the tips of the gastrosteges, and first row of scales. The only marking upon the head is a black postocular vitta, which extends along the upper borders of the labials and no farther, parallel to the commissure of the mouth. Pectus, throat, chin, and superior labials yellowish.” Measurements in mm.: Total length of largest known female specimen, 1980 mm.; tail, 340. Distribution: The typical form is known in Thailand from Chiang Mai province, and the Dong Rek mountains of central Thailand. Elsewhere it is known from Sikkim, Burma, Hong Kong and north- ern Indochina. It is probable that this typical form extends further south than my records show. Variation: The loreal may be twice as long as high; and of the seven to nine supralabials, two or three (rarely one) bordering the orbit. A presubocular is usually present. The scalerow-formula is usually, 23, 23, 19. In northern Indochinese territory this may be, 25 205 1! Elaphe radiata (Schlegel ) (Fig. 23) Coluber radiatus Schlegel, Essai sur la physionomie des reptiles vol. 2, 1837, p. 135, pl. 5, figs. 5-6; (type-locality, Java); Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, pp. 333-334; Catalogue of the snakes in the British Museum, vol. 2, 1894, pp. 61-62; Wall, Journ. Bombay Nat. Hist. Soc., vol. 19, 1910, p. 825; ibid., vol. 23, 1914, p. 206, fig.; ibid., vol. 29, 1923, p. 621; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 1, 1914, p. 95, pl.—. Caelognathus radiatus: Cochran, Proc. U. S. Nat. Mus., vol. 77, 1930, p. 6. Elaphe radiata: Pope, The reptiles of China, 1935, pp. 261-263, fig. 56; Bourret, Les serpents de Indochine, 1936, vol. 2, p. 211; Shaw, Shebbeare and Barker, Journ. Darjeeling Nat. Hist. Soc., vol. 14, 1939, p. 73; M. Smith, The fauna of British India including Ceylon and Burma . . . Reptilia and Amphibia; vol. 3, Serpentes; Dec. 1943, pp. 146-148, fig. 46. Coluber quadrifasciata Cantor, Proc. Zool. Soc. London, 1839, p. 51, (type- locality, Assam). Tropidonotus quinquefasciata Cantor, ibid., 1839, p. 54 (type-locality, Mergui, Tenasserim, Burma), 122, THe UNIVERSITY SCIENCE BULLETIN Diagnosis: Large snakes reaching a length of two meters; a. trans- verse black occipital bar making a median angle; scales in 19 rows at midbody; last labial below eye touching temporals; scales keeled; snout twice as long as eye; loreal longer than high; a median yellow or yellow-brown line bordered by a black stripe on each side. Description of species (from No. 33630, Nakhon Ratchasima province near Mauk Lek (Sara Buri province): Rostral about twice as wide as high, well visible above, posterior border convex; inter- nasals about as wide as long, the median common suture less than half as long as that between perfrontals; latter a little longer than wide; frontal longer than wide, slightly shorter than its distance from tip of snout; shorter than parietals; nostril between two nasals, posterior a little the larger; loreal longer than high; preocular reaches top of head but fails to form a suture with frontal; two postoculars, upper touching parietal; temporals, 2 + 2 + 3, the two anterior touching sixth labial; nine supralabials, fourth, fifth, and sixth bordering eye, eighth largest; mental small; ten infralabials, first five bordering first chinshields which are wider, but shorter than second pair. Scalerows: 21 on neck; 19 at mid-body, 17, before vent; anterior scales smooth; then at first, five median rows, posteriorly seven or nine rows, lightly keeled; scales with paired apical “pits.” Ventrals, 242: anal, single; subcaudals, 94. On sides a series of elongated black spots along third and fourth scalerows and a narrower series along outer scalerow and edge of ventrals; more posteriorly the upper row forms a continuous line. Color in life: Top of head olive becoming brown on temples and neck (top and side). A black transverse occipital bar connecting with eye by two black lines running forward; laterally bar reaching to level of ventrals and continued back as a narrow zig-zag black line for a short distance; a line below eye crossing supra- and infralabials; a diagonal black line from eye across back of mouth; a median clay-colored stripe, nearly three scales wide begins on occiput or near it, and runs back two thirds length of body; after a short distance this stripe, flanked by two broad black stripes vary- ing from two to three or four scales wide at intervals, gradually nar- rows posteriorly, becoming brown about middle of body and is finally lost in the ground color; posterior part of body nearly uniform brown, median part slightly lighter; tail light brown, unicolor on top and sides; chin and ventral surface of neck dirty ivory for about SERPENTS OF THAILAND Fic. 23. Elaphe radiata (Schlegel). No. 4049, Ratapoom, Songkhla province, southern Thailand. Actual total length, 542 mm.; tail, 99 mm. 724 Tue UNIVERSITY SCIENCE BULLETIN 16 centimeters; ventrals then begin to acquire gray pigment which becomes denser but lightens again near tail; subcaudal area uniform yellow-cream. Measurements in mm.; Total length, 1440; snout to vent, 1163; tail 277; width of head, 20; length of head, 30. Variation; Ventrals vary between 222 and 250, the subcaudals from 82 to 108. Rarely there are eight supralabials and occasionally there are but 19 scalerows around neck. There is some variation in the coloration. Occasionally the body may have a reddish cast, but the pattern of the markings is relatively constant. Distribution: The species occurs throughout Thailand in the low- lands, specimens having been taken in almost all areas where col- lections have been made. Outside Thailand the species ranges from Orissa and Sikkim in India—east and south to southern China, Indo-China, Malaya and the Indo-Australian Archipelago. Remarks: It feeds largely on small mammals frequenting the rice field. The juvenile figured is No. 404 from Ratapoom, Songkhla. Elaphe porphyracea (Cantor) Coluber porphyraceus Cantor, Proc. Zool. Soc. London, 1839, p. 51 (type- locality, Mishmi Hills, “Asam” Abor country ). Two subspecies are recognized, the typical one entering our territory in the northwestern part. Elaphe porphyracea porphyracea (Cantor) (Fig. 24) Coluber prophyraceus Cantor, Proc. Zool. Soc. London, 1839, p. 51 (type- locality, Mishmi Hills, Assam, Abor country); Giinther, The reptiles of British India, 1864, p. 239, pl. 20, fig. J; Boulenger, Catalogue of the snakes in the British Museum, vol. 2, 1894, p. 34; Annadale, Journ. Asiat. Soc. Bengal, 1905, p. 175; Wall, Journ. Bombay Nat. Hist. Soc., vol. 18, 1908, p. 326; ibid., vol. 19, 1909-1910, pp. 345, 827; ibid., vol. 29, 1923, p. 620; ibid., vol. 30, 1925, p. 812; Rendahl, Ark. Zool. K. Sven. Vet. Akad. Stockholm, vol. 29, A, 1937, p. 16. Ablabes prophyraceus: Boulenger, The fauna of British India . . . Rep- tilia and Batrachia, 1890, p. 308; Sclater, List of the snakes in the Indian Mus., 1891, p. 19; Wall and Evans, Journ. Bombay Nat. Hist. Soc., vol. 13, 1901, p. 611; Venning, Journ. Bombay Nat. Hist. Soc., vol. 20, 1910, p. 337; Annandale, Rec. Ind. Mus., 1911, p. 217; ibid., 1912, pp. 37, 47, 53. Elaphe porphyracea: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 48; Pope, The reptiles of China, 1935, pp. 253-257, fig. 55: Bourret, Les serpents de l’Indo-Chine, 1936, p. 187; Shaw, Shebbeare, and Barker, Journ. Darjeeling Nat. Hist. Soc., vol. 14, 1939, p. 72: M. Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 154-156, fig. 47. SERPENTS OF THAILAND 725 Elaphe porphyracea prophyracea: M. Smith, Rec. Ind. Mus., vol. 42, 1940, p. 480. Coronella callicephalus Gray, Ann. Mag. Nat. Hist., ser. 2, vol. 12, 1853, p. 390 (type-locality, Khasi Hills). Fic. 24. Elaphe porphyracea porphyracea (Cantor). No. 36526, Doi Suthep, near Chiang Mai, Chiang Mai province, northem Thailand. Actual total length, 307 mm.; tail, 51 mm. 726 THE UNIVERSITY SCIENCE BULLETIN Diagnosis: Somewhat exceeding a meter in length (1100 mm. ); scales smooth in 19 rows; loreal present; above reddish to magenta with a series of about 14 to 16 darker bands on body; two lines from first band extended forward along sides of head to eye; a median stripe on head; two dorsolateral black lines on posterior third of body and tail, a total of four darker bands indicated on tail. Description of species (from No. 20926, Mae Hong Son): Rostral nearly one-half wider than high, well visible above; internasals sub- quadrangular, about half area of prefrontals; frontal as long as its distance from tip of snout, little shorter than parietals; parietals bordered behind by four scales; nostril between two nasals, pos- terior one largest; loreal present, considerably longer than high; preocular large, reaching top of head but separated from frontal; two postoculars, both touching parietal; temporals, 1 + 2, anterior touching only lower postocular; eight supralabials, fourth and fifth border eye; ten infralabials, first five border first chinshieds which are larger than second pair. Scale-formula: 23 (occiput), 19, 19, 17. Scales smooth through- out. Ventrals, 214; subcaudals, 66; anal single. Maxillary teeth, 20-21, anterior ones largest. Color in life: Above dull red to magenta with dark-edged cross- bars, centers of which are practically same shade as body inter- vening areas double or triple length of cross-bars; two narrow black lines extend straight forward to eyes from first bar; median black line on head from anterior edge of frontal to end of parietals; black edges of cross-bars bordered with lighter color at least an- teriorly. Ventrals and subcaudals ivory-white except on outer edges; chin ivory-white; supralabials and top of head grayish brown; a pair of dorsolateral black lines begin at about 12th cross-bar, extending backwards. Measurements in mm. (Nos. 20926 and 36526): Total length, 983, 307; tail, 156, 51; width of head, 16, 6.7; length of head, 28.5, 14.4. Variation: The young specimen, No. 36526, is from Doi Suthep, collected by O. Gordon Young. The bands are coal black, 14 in number on body, and four on tail; the dorsolateral bands begin on the twelth bar. There are nine supralabials (the third being divided probably abnormally, thus the fifth and sixth border eye). The second pair of chinshields are transversely divided leaving three pairs of chin- SERPENTS OF THAILAND Me shields. Ten lower labials present, the median stripe on head reaching to the internasals. The first temporal touches neither of the postoculars but both touch the parietal. The ventrals are 208, subcaudals, 69; the anal is divided. The known ventral range is 190 to 218; the subcaudal range, 52 to 76. Distribution: The snake has been taken only in northern Thailand in mountains. It is known from the provinces of Mae Hong Son, and Chiang Mai. Outside of Thailand the species occurs in the Eastern Himalayas, Assam, Burma, Yunnan, west China, northern Malaya, and Sumatra. No specimens have been reported between Doi Suthep, Chiang Mai and the Cameron Highlands (4500 ft. elev.) in northern Malaya. Elaphe flavolineata (Schlegel) (Fig. 25) Coluber flavolineata Reinwardt, Bull. Sci. vol. 9, 1826, p. 237 (nom. nud.); Schlegel, Essai sur la physionomie des serpents, vol. 2, 1837, p. 141 (type- locality, Java); Stejneger, Nyt. Mag. Naturvid. Kristiana, vol. 60, 1922, (2), p. 78. Coluber melanurus (non Shaw 1802) Schlegel, Essai sur la physionomie des serpents, vol. 2, 1837, p. 141, pl. 5, figs. 11, 12 (type-locality Java); Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 334; Catalogue of the snakes in the British Museum, vol. 2, 1894, pp. 60-61; Annandale, Journ. Asiat. Soc. Bengal, new series, vol. 1, 1905, p. 173; M. Smith, Journ. Nat. Hist. Soc. Siam., vol. 2, no. 2, 1916, pp. 160-161; Wall, Journ. Bombay Soc. Nat. Hist., vol. 29, 1923, p. 261. M. Smith, Bull. Raffles Mus., no. 3, 1930 p. 48. Composoma melanurum: Ginther, The reptiles of British India, 1864, pp. 244- 245. Elaphe melanurus: Barbour, Mem. Mus. Comp. Zool., Harvard Coll. vol. 44, 1912, p. 117; Sworder, Singapore Nat., no. 2, 1922, p. 63, (Singapore). Elaphe flavolineata: M. Smith, Bull. Raffles Mus., no. 3, Apr. 1930, p. 48; The fauna of British India including Ceylon and Burma . . . Reptilia and Amphibia, vol. 8, Serpentes, 1943, pp. 148-149. Diagnosis: A large snake reaching approximately one-and-one- half meters in length; color black and yellowish brown; loreal not particularly small; the last labial touching eye also in contact with temporal scales; scales on posterior two thirds of body strongly keeled; scales in 19 rows; ventrolateral spots on anterior part of body; two dark spots radiating below eye; no transverse bar on neck; posterior part of body and tail (above and below) black; a pair of dorsolateral black lines on anterior part of body. Description of species (from EHT-HMS No. 9. Kuala Lumpur, Malaya): Rostral one-and-one-third times as wide as high, very narrowly visible above; internasals as wide as long, much smaller 728 THE UNIVERSITY SCIENCE BULLETIN than pretrontais, little wider than long; frontal scarcely angulate anteriorly, nearly triangular, its length equal to its distance from rostral, equal to common parietal suture but shorter than parietal scutes; nasal divided, posterior part largest; loreal longer than high; one large preocular reaching upper side of head but not touching frontal; two postoculars; temporals, 2 + 2, both anterior temporals touching sixth labial; supralabials, 9-9, fourth, fifth, and sixth bordering orbit; mental small, its labial border much less than that of rostral; infralabials, 10-11, six touching first pair of chinshields which are larger, but approximately same length as second pair. Scale-formula: 21, 19, 17; anteriorly 11 rows keeled, 17 rows more posteriorly; outermost row unkeeled throughout. Ventrals, 227, somewhat angulate laterally; anal, single, subcaudals, 104. Color: Head above and on sides uniform olive-brown; supra- labials yellow-ivory except along their upper edges; a blackish bar below eye extending across infralabials; a diagonal black bar from eye across supralabials terminating on the eighth; a black stripe from posterior temporals to first ventrolateral black spot on ninth and tenth ventrals; at level of twentieth ventral, a pair of dorsal black stripes begin, separated mesially by three scalrows, the stripes two scalerows wide; these continue back to about level of the one hundredth ventral; area between black stripes yellow-brown; from the termination of the lines the brown of back gradually darkens to black continuing so to end of tail. Anteriorly on outer edge of ventrals a series of twelve black spots, often extending onto first scalerow; above these on side a lighter area, then a short black mark with another lighter area between this and the black dorsal stripe. The light areas associated with black spots extend onto the skin between scales. Measurements in mm.; Snout to vent, 1095; tail, 324; total length, 1419; width of head, 17; length of head, 31.5. Variation: Ventrals vary between 193 and 234; subcaudals, 89 and 115. The young may have ocelli with bright yellow centers as- sociated with the black spots. The scales about neck may be in 21 or 19 rows. Distribution: In Thailand the species has been found only at “Bangnara’ [= Narathiwat], in the province of that name. : The species is widespread occurring in the Indo-Australian Archi- pelago, Malaya, Nicobar Island, Andaman Islands and Burma. I have not found it in Thailand, and I use a Malayan specimen for the description. SERPENTS OF THAILAND Fic. 25. Elaphe flavolineata (Schlegel). EHT-HMS No. 9, Kuala Lumpur, Selangor, Malaya. Actual total length, 1419 mm.; tail, 324. 730 THE UNIVERSITY SCIENCE BULLETIN Genus Ptyas Fitzinger Ptyas Fitzinger, Systema Reptilium, 1843, p. 26 (type of genus, blumenbachii). Zamenis Gagler, Natiirliches System der Amphibien 1830, p. 188 (type of genus Aesculapii Lacépéde [non Linnaeus] “Aescaulapechlange’” ). Dignosis: Large snakes, with head distinct from the more or less cylindrical body; eye large to very large, the pupil round; scales in 17-15 rows (more rarely 18, 16, 14); subcaudals, paired; apical pits present; normally two or three loreals; a presubocular usually pres- ent. Maxillary teeth, 20 to 28, increasing slightly in length and thick- ness posteriorly. Three species occur in Thailand. They are large, conspicuous snakes that move very rapidly on the earth, and at least one is a good tree climber. They may be separated by the following key: Kry TO SPECIES OF Ptyas IN THAILAND 1. Number of scalerows never in even numbers; scales in 15 rows at middie of body; ventrals less numerous, 160-187 subcaudals, 120-147; young with white bands; adults showing lineation in posterior part of body..... korros Scalerows often even or at least so in posterior part of body........... 2 . Seales in 17 or 16 rows at middle of body; ventrals more numerous 190- 213, subeaudals, 100-146; darker bands at least in posterior part of the body, that may be somewhat reticulated or not; labials and neck scales with: blackisuttures 0037 aspen es ee ee MUCOSUS Seales only in even number of rows, 16 or 14 rows at middle of body, the median rows (four or six) keeled; ventrals, 200-215, subcaudals, 110-118, yellowish brown on posterior part of body with six black irregular longi- tudinal stripes tending to form a reticulation................. carinatus to Ptyas korros (Schlegel ) (Figs 26, 27) Coluber korros Schlegel, Essai sur la physionomie des serpents, vol. 2, 1837, p 139; Abbildungen . . . Amphibien, 1840, p. 99, pls. 27-28, figs. 1-6 (type-locality, Java); Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 121. Ptyas korros: Giinther, The reptiles of British India, 1864, p. 250; Theobald, Journ. Linn. Soc., Zoology, vol. 10, no. 41, May, 30, 1868, p. 46. Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, 1885, pp. 54, 65; Sclater, Hand list of the snakes in the Indian Museum, 1891, p. 26; M. Smith, Journ. Nat. Hist. SOC: Siam, vol. 1, 1914, p. 94, 187; ibid, vol. 6 pt. 38, 1915, p. 213; ibid, vol. 2, 1916, p. 16 (Pattani, Nakhon Si Thammarat); SS crdce Singapore Nat., oa 2, 1922, p. 62; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 6, no. 2, 1923, D. 202; Kopstein, Treubia, vol. 11, 1930, p. 301, fig.; M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 47; Pope, The reptiles of China, 1935, pp. 217-220; Taylor, Proc. Acad. Nat. Hist. Soc. Philadelphia, vol. 86, 1936, p. 305 (Chantaboon = Chanthaburi, Chanthaburi province ); Bourret, Les serpents de l’Indo-Chine, 1936, p. 176; Shaw, Shebbeare, and Barker, Journ. Darjeeling Nat. Hist. Soc., vol. 14, 1939, p. 71; M. Smith, The Fauna of British India, Ceylon and Burma . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 162-163; Tweedie, The snakes of Malaya, 1954, p. 48, fig. 8 (Second Ed., 1957); Taylor and SERPENTS OF THAILAND 731 Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, p. 1136 (Chaiyaphum and Kalasin provinces). Zamenis korros: Boulenger, The fauna of British India, Ceylon, and Burma, Reptilia and Batrachia, 1890, p. 324; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, pp. 384-385; ibid, vol. 3, 1896, p. 621, in Addenda; Flower, Proc. Zool. Soc. London, 1899, p. 666; Wall and Evans, Journ. Bombay Nat. Hist. Soc., vol. 13, 1900-1901, p. 353, 620; Wall, ibid, vol 18, 1908, p. 326; ibid, vol. 19, 1909, pp. 345, 622; Boulenger, A. verte- brate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 137; Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 618. Ptyas korros chinensis Mell, Sitz. Ber. Ges. Nat. Freunde, Berlin, 1930, p. 320 (type-locality, Yao-shan, Kwangsi, China). Ptyas korros indicus Mell, Lingnan Sci. Journ., vol. 8, p. 208 (type-locality SW Yunnan, China). Liopeltis libertatis Barbour, Proc. Biol. Soc. Washington, vol. 23, p. 169 (type- locality, Buitenzorg, Java); Dunn, Amer. Mus. Novitates, No. 287, 1927, p. 1. Diagnosis: Large snakes reaching a length of at least two meters; young gray on anterior half of body with dotted blue-white trans- verse bars, tending to disappear in older individuals which become nearly uniform brown or gray-brown anteriorly; scales may be slightly edged with white; farther back this white portion widens forming eleven pale longitudinal lines, separated by brownish or grayish lines. Three (two) loreals; one presubocular; scalerows: 15, 15, 11; ventrals, 160-187. Description of species (from No. 34162 2 Bangkok): Rostral wider than high, narrowly visible above; internasals about two thirds area of prefrontals; latter scales touching two (one) loreals laterally; frontal longer than wide, longer than its distance from end of snout, as long as parietals; nasal divided, anterior portion the largest, anterior edge of posterior part deflected downward; three (two) loreals; large preocular reaching top of head but not touching frontal; supraoculars forming a slight shelf above eye and preocular region; small presubocular wedged between third and fourth supralabials; two postoculars, upper largest; temporals, 2+ 2-+ 2, upper anterior touching both postoculars; supralabials eight, fourth and fifth bordering eye; infralabials ten, first five touching first chinshields which are smaller than second pair; latter scales touch each other anteriorly, separated by two scales pos- teriorly; eye large, its length equal to its distance from nostril. Scale-formula: 20 (occiput), 15, 15, 11, scales smooth. Ventrals, 170; subcaudals, 104; anal, divided. Color: Head and anterior part of body gray-brown (bluish gray when epidermis is shed); scales with grayish-white edges, at first 732 THe UNIVERSITY SCIENCE BULLETIN Fic. 26. Ptyas korros (Schlegel). No. 33310 young, Muak Lek, Sari Buri province, central Thailand. Actual total length, 593 mm.; tail, 211 mm. scarcely noticable but farther back light edges of two scalerows make pale lines that widen posteriorly and are separated by lines of brownish ground color; farther back this dark color becomes black, concentrated chiefly in a single median posterior spot on the scale with a slightly darker edging; major part of each scale on posterior fifth or sixth of body and on tail, ivory-white; labials, chin, and ventral surface brownish white to nearly ivory; white posteriorly in subcaudal region; sides and tip of snout whitish. Measurements in mm.: Total length, 1402; tail, 443; width of head, 19; length of head, 35. Distribution: The species occurs throughout Thailand in low- lands. It has been taken in the provinces of Yala, Trang, Nar- athiwat, Pattani, Phattalung, Nakhon Si Thammarat, Bangkok, Sara Buri, Khorat, Chaiyaphum, Kalasin, Chiang Mai and Chon Buri. Outside of Thailand the species ranges in Assam and Burma, Laos, South China, Cambodia, Malaya, Sumatra, Java, and Borneo. SERPENTS OF THAILAND 733 Fic. 27. Ptyas korros (Schlegel). No. 34162, Bangkok (city). Actual total length, 1402 mm.; tail, 443 mm. Remarks: This large snake is one that is frequently mistaken for the King Cobra. Needless to remark, the species is entirely harm- less, feeding largely on frogs, rats, and other small rodents and occasionally on other reptiles and birds. They are often seen in bushes and low trees especially in the neighborhood of water. They remain quiet and one may approach quite close to them if the approach is quiet and slow. When in grass the head may be raised thus looking even more like the Hamadryad. The young have numerous narrow whitish bars and look very different from the adults. The species lays eggs, usually not more than twelve. When hatched, the young measure about 375 millimeters in length. 734 THe UNIVERSITY SCIENCE BULLETIN Ptyas mucosus (Linnaeus ) (Fig. 28) Coluber mucosus Linnaeus, Museum Adolphi Friderici Regis, vol. 1, 1754, p. 37, pl. 23; Systema naturae, ed. 10, 1758, p. 226 (type-locality, India ) based on Russell, An account of Indian serpents collected on the coast of Coromandel, vol. 1, 1796, p. 40, pl. 34; Andersson, K. Sven. Vet-Akad. Handl. Stockholm, vol. 24, 1899, p. 25. Zaocys mucosus: Wall, Journ. Bombay Nat. Hist. Soc., vol. 23, p. 168; ibid., vol. 26, p. 566; Spolia Zeylanica, 1921, p. 399. Ptyas mucosus: Ginther, The reptiles of British India, 1864, p. 249; Theobald, Journ, Linn. Soc. vol. 10, no. 41, Zoology, May 30, 1868, p. 46; Blanford, Proc. Zool. Soc. London, 1881, p. 221; Tirant, Notes sur les reptiles et les Batraciens de la Cochinchine et du Cambodge, 1885, pp. 54, 55, 65; Wall, Ophidia Taprobanica or the snakes of Ceylon, 1921, pp. 172-190, fig. 40; Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 617; Prater, Journ. Bombay Nat. Hist. Soc., vol. 30, (1), 1924, p. 169; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 6, no. 2. Oct. 1923, p. 202 (Hainan); M. Smith, Bull. Raffles Mus. no. 3, 1930, p. 47; Subrahmaniam, Journ. Bombay Nat. Hist. Soc., vol. 37, 1934, p. 743; Taylor, Proc. Acad. Nat. Sci. Philadelphia, vol. 86, 1934, pp. 304-305 (Chiang Mai); Pope, The reptiles of China, 1935, pp. 220-223; Fraser, Journ. Bombay Nat. Hist. Soc., vol. 29, 1937, p. 475; Shaw, Shebbeare and Barker, Journ. Darjeeling Nat. Hist. Soc., vol. 14, 1939, p. 68; Taylor, Univ. Kansas Sci. Bull., vol. 33, pt. 2, Mar. 20, 1950, pp. 543, 544 pl. 15, fig. 1; Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, p. 11385 (Nakhon Pathom province). Zamenis mucosus: Boulenger, The fauna of British India, including Ceylon and Burma, Reptilia and Batrachia, 1890, p. 324; Catalogue of the snakes in the Bristish Museum, 2nd Ed., vol. 1, 1893, pp. 385-386; Ferguson, Journ. Bombay Nat. Hist. Soc., vol. 10, 1895, p. 71; Beadon, ibid., vol. 20, 1910, p. 228; Millard, ibid., vol. 17, 1906, p. 245; Venning, ibid., vol. 20, 1910, p. 339; Millet, ibid., vol. 19, 1909, p. 758, 759; Fenton, ibid., vol. 19, 1910, p. 1002; Wall, Journ. Bombay Nat. Hist. Soc., vol. 17, 1906, p. 259, pl. 3: ibid., 17, 1907, p. 259, ibid., vol. 18, 1907, p. 113; ibid., vol. 19, 1909, p. 622; ibid., vol. 21, 1911, p. 134; Boulenger, A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, pp. 137-138; Nikolski, Faune de la Russie, vol. 2, 1916, p. 79; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917. p. 98; McCann, Journ. Bombay, Nat. Hist. Soc., vol. 38, 1935, p. 409; Taylor, Proc. Acad. ‘Nat. Sci. Philadelphia, vol. 86, 1934, pp. 304-305 (Chiang Mai prov.); Bourret, Les serpents de lIndochine, 1936, p. 178. Zaocys mucosus: Wall, Journ. Bombay Nat. Hist. Soc., vol. 23, 1914, p. 168; ibid., vol. 26, 1919, p. 566. Coluber blumenbachii Merrem, Tentamen systematis amphibiorum, 1820, p. 119 (type-locality, Bengal). Coluber dhumna Cantor, Proc. Zool. Soc. London, 1839, p. 51 (type-locality, “Carnatic Orissa, Bengal, Nepal, Asam, Arracan, Tenasserim”’ ). Leptophis trifrenatus Hallowell, Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 503 (type-locality, Hong Kong). Diagnosis; Large snakes bearing a superficial resemblance to the king cobra and to the large rat snake Ptyas carinatus; “reaching a length of 11 ft. 9 in.” Scales usually in 17 (or 16) rows; ventrals, 190-213; subcaudals, 121-146. Maxillary teeth, 20-25, more or less subequal; sutures of supralabials and infralabials blackish, with numerous blackish marks SERPENTS OF THAILAND 735 Fic. 28. Ptyas mucosus (Linnaeus). No. 17, Bangkok (city), Thailand. Actual total length, 1702 mm.; tail, 439.5 mm. on chin and forepart of neck (ventral); narrow irregular dark-brown bands on posterior part of body. Description of species (from No. 34165, Bangkok): Rostral wider than high, narrowly visible above; internasals longer than wide, of about half area of prefrontals; latter bordering two loreals, wider than long, the sides bent down; frontal as long as wide, distinctly shorter than its distance from rostral; only very slightly shorter than parietals; latter scales about a fourth longer than wide; nasal divided, anterior part largest, posterior part with its anterior edge deflected 736 THE UNIVERSITY SCIENCE BULLETIN downwards; three small loreals; two preoculars, upper large, its upper part reaching top of head and with supraoculars forming a small projecting shelf above and in front of eye; two postoculars; temporals, 2 + 2; supralabials, 8-9, the fourth and fifth bordering eye; ten infralabials, five bordering first pair of chinshields which are distinctly shorter than second pair; latter in contact anteriorly, separated by one or two small scales posteriorly. Scale-formula: 24 (occiput), 21, 17(16), 14, smooth with paired apical pits. Ventrals, 197; subcaudals, 117; anal, divided. Color; Head dark brown above (with epidermis removed, fawn to gray); snout yellowish; labials yellowish with black sutures on both supra- and infralabials; scales on neck and anteriormost ventrals similarly marked; body amber to brown, latter three fourths of body with narrow irregular bands or alternating broken bands on top and sides of body, passing on to venter but not crossing it (sometimes forming a reticulum above); numerous brownish spots on underside of posterior ventrals and subcaudals; venter light yellowish brown anteriorly, growing more yellowish posteriorly; subcaudals yellow- ish ivory. Measurements in mm.; Total length, 1753; tail, 463; width of head, 28; length of head, 47. Distribution: Found widely distributed in Thailand. Outside of Thailand the range extends from Transcaspia east through India, Ceylon to Indochina, and Malaya. Remarks: This species is somewhat less common than Ptyas korros, and when specimens are found with the scalerows 16 and 14 it is difficult to distinguish them from Ptyas carinatus which likewise has an even number of scalerows and three loreals. Moreover, mucosus may have the median scales keeled. Under these circum- stances the color differences will serve to distinguish them. Ptyas carinatus (Gunther ) Coryphodon carinatus Giinther, Catalogue of the colubrine snakes in the collec- tion of the British Museum, 1858, p. 112 (type-locality, Borneo). Zaocys carinatus: Gimther, The reptiles of British India, 1864, p. 256 (foot- note “Our largest example is nearly 10 feet long, the tail being one-fourth” ) ; Boulenger, Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, p. 377, pl. 27, fig. 1; Ann. Mus, Civ. Genova, ser. 2, vol. 13, 1893, p. 324; A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, pp. 135-136; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 160 (“Patani,” Nakhon Si Thammarat, “Northern Siam”); de Rooji, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, p. 73 (Malay Peninsula, Sumatra, Java, Borneo); Sworder, Singapore SERPENTS OF THAILAND Wt Nat., no. 2, 1922, p. 62, (Singapore); Joynson, Journ. Nat. Hist. Soc. Siam, vol. 6, 1927, p. 314; M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 46; Bourret, Les serpents de l’Indochine, 1936, p. 173, fig.; M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 164-165. Ptyas carinatus Wall, Journ. Bombay Nat. Hist. Soc., vol. 31, 1926, p. 562. Diagnosis: Maxillary teeth, 22-26; internasals two thirds of the length of prefrontals; two to four loreals; eight to ten supralabials, fourth and fifth or fifth and sixth bordering orbit. Scale-formula: 18, 16 (14), 12; four to six median scalerows keeled. Description of species: Rostral broader than high narrowly visi- ble from above; internasals shorter than prefrontals; frontal some- what longer than broad about as long as its distance from tip of snout, shorter than parietals; parietals a little longer than broad; supraoculars about as wide as frontal; nasal divided, anterior part a little the larger; loreals, two (three or four rarely); a preocular reaching top of head but not touching frontal; two postoculars; temporals, 2 -+ 2; supralabials nine, fifth and sixth border orbit; infralabials eleven, five bordering anterior chinshields which are distinctly shorter than posterior. Scalerows: 18, 16 (14), 12, four to six median scalerows with keels: ventrals, 208-215; subcaudals, 109-117; anal, divided. Color: Olive-brown to blackish on anterior part of body with more or less distinct cross-bands and yellowish-fawn spots; gradually ground color becomes buff-brown or yellowish brown posteriorly with a large black network; underside of head and venter yellowish anteriorly, black and yellowish posteriorly. Measurements in mm.: Total length, 3000; tail, 750. Variation: The ventral and subcaudals show rather a limited amount of variation; for the ventrals, 208-215; for subcaudals, 110-118. Supralabials vary from eight to ten, and the number of loreals may be two, three, or four. Distribution: The species occurs widely in Thailand, specimens being known from Lampang (Me Pow Forest, 20 mi. E. of Muang Ngow ); Chiang Mai, Nakhon Si Thammarat, Pattani (or Yala). Outside of Thailand the species ranges through Malaya, Singapore, Borneo, Java, Sumatra, to Palawan (Philippines). Remarks: This is not a rare snake but many specimens taken fail to preserve well because of their size and consequently are discarded. The snake is used much for food. 29—1367 738 THe UNIVERSITY SCIENCE BULLETIN This species has long been retained in the genus Zaocys. The definitions of the two genera Ptyas and Zaocys are very similar and the similarity between Ptyas korros, Ptyas mucosus, and Zaocys carinatus is such that it seems well to place them together. The chief difference in the genera was presumed to be that Zaocys had an even number of scalerows, and Ptyas an uneven number. It has been found that Ptyas mucosus very frequently may have the scalerows in even numbers throughout a part or even over very much of the body. In Ptyas korros there is an odd number of scalerows. Thus it would appear that Ptyas mucosus breaks down the es- sential distinction that would warrant separation and a recognition of two genera. In consequence I follow Wall, loc. cit. in making Zaocys a synonym of Ptyas, the older name. Malcolm Smith has maintained them as separate genera. Genus Lycodon Boie Lycodon Boie, in Ferussac’s, Bull. Sci. Nat., vol. 9, 1826, p. 238 (part.); Fitzinger, Neue Classification der Reptilien, 1826, pp. 29, 30 (type of genus, Lycodon aulicus). M. Smith, Fauna British India . . . Ser- pentes, 1943, pp. 255-257, fig. 88. Ophites Wagler, Natiirliches System der Amphibien, 1830, p. 186, (type of genus, O. subcinctus). Diagnosis: Small snakes not reaching one meter in length (ex- cept subcinctus). Head distinctly flattened, slightly distinct from body; eye with vertically elliptic pupil; body rather short; scales usually in 17 rows (one species 15, one 19 rows), with apical pits; some forms with lateral keels on ventrals; subcaudals (except in one Indian species) paired. Hypapophyses absent on posterior vertebrae. Maxillary arched with three or four small teeth pre- ceding two enlarged teeth, then a curving diastema followed by 7 to 15 teeth, of which the last two are enlarged; nostril between two nasals, or nasal only partially divided; loreal in contact with internasal or not, the scale elongated; three supralabials, third, fourth, and fifth bordering eye; anal normally divided. The genus is distributed from Transcaspia throughout the Oriental Region, south through Malaya and the Indo-Australian Archipelago. Approximately 20 species are known, four of which occur in Thailand. They may be differentiated by the following key: Key To Tuat Species oF Lycodon 1. No preocular present; prefrontal and loreal border eye; scales feebly keeled; one anterior temporal; posterior nasal higher than anterior; usually eight supralabials: ventrals 197-230 52 2 he eee s. subcinctus A preocular separating prefrontal from eye; loreal separated from border of eye, or not; scales smooth or keeled. SERPENTS OF THAILAND 739 2. Snout not projecting beyond lower jaw; loreal usually just touching or sep- arated from internasal: two anterior temporals; nine supralabials; first chinshields two to three times size of second pair................... laoensis Snout projecting beyond lower jaw. 3. Seales smooth; ventrals strongly angulate; nasals subequal; loreal in con- tact with internasal, not touching eye; a white or yellow nuchal band, speckled with lavender; body reticulated with lavender and fawn . .capucinus Median scalerows feebly keeled; ventrals feebly angulate; posterior nasal larger than anterior; loreal borders eye, well separated from internasal; no white or yellow nuchal band; numerous transverse light and dark bands, the dark ones becoming narrower posteriorly....................-. fasciatus Lycodon subcinctus subcinctus Boie Lycodon subcinctus Boie, Isis, 1827, p. 551 (Based on Russel, Indian Serpents, vol. 2, 1801, p. 44, pl. 41, (type-locality, Java): Boulenger, Catalogue of the snakes in the British Mus., vol. 1, 1893, pp. 359-360; Flower, Proc. Zool. Soc. London, 1899, p. 665; Boulenger, A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, pp. 133-134; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, pp. 108, 109, fig. 45: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 160 (Pattani); Sworder, Singapore Nat., no. 2, 1922, p. 62; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 6, 1923, p. 202; M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 46 (Gunong Tahan, Malaya, 5400 ft. ). Ophites subcinctus: Giinther, The reptiles of British India, 1864, p. 322; Blan- ford, Proc. Zool. Soc. London, 1881, p. 222, pl. 21, fig. 2; Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, 1885, p. 46: Bourret, Les serpents de I’Indo-chine, 1936, p. 157; Pope, Reptiles of China, 1935, pp. 196-197, fig. 45 (head); Herklots, Hong Kong Nat., vol. 6, 1935, p. 195, fig. (head). Elapoides annulatus Sauvage, Bull. Soc. Philom., ser. 7, vol. 8, p. 144 (type- locality, Sumatra). Lycodon platurinus Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 916. Diagnosis: Scales in 17 rows; no preocular present, permitting both prefrontal and loreal to border eye; scales feebly keeled; young with 10-13 broad purplish-black bands separated by white cross- bars: white bars lost in posterior part of body in adult specimens, and only a trace of the white remains in old specimens. Reaching a length of one meter. Description of species (from No. 35656, Tonka Harbor Tin mine, Ronpibon, Nakhon Si Thammarat): Snout broad, depressed; rostral about a third wider than high, very slightly visible above; internasals less than half area of prefrontals, the suture between them half that between prefrontals; latter scales about as wide as long, bordering eye; frontal wider than long, its length equal to or slightly shorter than prefrontals, more than twice width of a supra- ocular; parietals elongate, their length only a little less than their distance from rostral; nasal divided, anterior part quadrangular, the posterior part larger, subtriangular; nostril large; loreal nearly twice as long as high bordering eye, touching three supralabials, but separated widely from internasal; no preocular, two postoculars; temporal formula: 1+ 2-+ 3; eight supralabials, fourth and fifth 740 THE UNIVERSITY SCIENCE BULLETIN (on left side, third, fourth, and fifth) bordering eye; eye with a vertically oval pupil; infralabials nine, four bordering the anterior pair of chinshields which are much larger than second pair. Scale-formula: 20 (occiput), 17, 17, 15, the median ten rows with weak keels. Ventrals, 216, laterally angulate with a slight notch on each side; anal, divided; subcaudals, 83. Color in life: Dark purplish-black anteriorly, growing somewhat lighter posteriorly; a cream-ivory spot on sides of occiput back of jaw-angle; at level of 23rd ventral an ivory-white band, widest on venter, dorsal portion bearing many blackish scales; second light band begins on a level with fifty-first ventral, represented only by a few scattered ivory spots: third light band is barely indicated by some light flecks on venter with a small lateral spot on one side. Chin and throat grayish white; most ventrals grayish, color more intense on anterior part of each scale; subcaudal region darker gray than venter. Measurements in mm.: Total length, 923; tail, 173; width of head, 14.3; length of head, 23. Variation: The typical color pattern of the young consists of 10 to 13 light crossbands surrounding the body, the scales sometimes edged with black; the hinder part of the head white. In specimens from the northern part of the range the white bands are speckled with black and the underside of the tail may be white. The ventrals vary between 197-230,* the subcaudals 71-90; the anal scale is oc- casionally single. Rarely three postoculars are present. Distribution: The species while known from a few localities in Thailand presumably ranges over the whole country. Outside of Thailand it is known from Southern China, Laos, Viet Nam, Cam- bodia, the Malay Peninsula, and the Indo-Australian Archipelago to the Philippines. Remarks: 1 captured the described specimen within the hollowed base of a forest tree, ensconsed behind a loose fragment of wood at Ronpibon, Nakhon Si Thammarat. It moved with considerable rapidity. A second specimen was taken at Na Pradoo, Pattani. Lycodon laoensis Gimnther (Figs. 29, 30, 31) Lycodon laoensis Giinther, Reptiles of British India, 1864, pp. 317-318, (type- locality, Laos, French Indo-China); Boulenger, Catalogue of the snakes in the British Museum, vol. 1, 1893, p. 354, pl. 24, fig. 2; Laidlaw, Proc. Zool. Soc. London, vol. 2, 1901, p. 576; Boulenger, A vertebrate fauna of the Malay Peninsula from the Isthmus of Kra to Singapore, Reptilia and * De Rooij states 128-230 and 61-90 for subcaudals. I suspect 128 is a typographical error. SERPENTS OF THAILAND TAL Batrachia, 1912, p. 132; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 160; Bull. Raffles Mus., no. 3, 1930, p. 46; The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Ser- pentes, Dec. 1943, pp. 259-260. Diagnosis: A small snake (475 mm.); body bluish, or brownish black, with narrow bright yellow or white cross-bars which expand laterally, the expansions enclosing triangular or other spots of same Fic. 29. Lycodon laoensis Ginther. No. 333084, one and _ one-half miles west of Muak Lek, Sara Buri, but actually in the province of Nakhon Ratchasima, Thailand. Actual total length, 445 mm.; tail, 103 mm. 742 THe UNIVERSITY SCIENCE BULLETIN body color; a preocular separating prefrontal from eye, in contact with frontal; loreal elongate, not entering orbit, slightly in contact or not with internasal; anterior and posterior parts of nasal equal or anterior largest. Maxillary teeth in two groups separated by a diastema. A single apical pit on each scale. Description of species (from No. 33308 ¢ one-and-one-half miles west of Muak Lek, Sara Buri, but actually in Nakhon Ratchasima province): Head moderately distinct from neck, rather flattened, the temporal regions not elevated; rostral much broader than high, part visible above less than half length of prefrontals; internasals less than half area of prefrontals, about as wide as long, their mutual suture about half as long as that between prefrontals; latter scales longer than wide; frontal short, four-fifths as wide as long, shorter than its distance from tip of snout, and much shorter (one third or more) than parietal; supraocular not more than half width of frontal. Nasal divided, anterior part a little larger than posterior; an elongate loreal narrowly touching internasal, but separated from eye by preocular and third labial; preocular large, in contact with 1 the frontal; two postoculars; temporals, — + 4 + 3; ten supralabials, 9 third, fourth, and fifth entering orbit; ten infralabials, five touch- ing first chinshields which are about three times size of second pair; first pair of infralabials relatively large. Scale-formula: 24 (occiput), 17, 17, 15; scales-smooth, with a single minute apical pit. Ventrals, 166; subcaudals, 71; anal, di- vided; ventrals more or less distinctly angulate. Maxillary teeth: Anterior group of 5, first three small, then two much enlarged; after a curving diastema, eight small teeth, fol- lowed by two enlarged teeth that are smaller than the anterior enlarged ones. Color in life: Above grayish blue, scales a little darker on edges. Top of head ultramarine, the color terminating in a point at ter- minus of suture between parietals; anterior supralabials same color; posterior labials cream, upper edges of labials bordering orbit, pigmented; a broad yellow band across occiput narrowed mesially. Chin and entire ventral surface yellowish ivory; subcaudal region, gray-white. A series of bright yellow transverse bars cross back, widening laterally and enclosing somewhat triangular spots of dark color; these are as wide as three transverse scalerows anteriorly but narrow posteriorly to width of one row (these without lateral spots ). The spacing of dark blotches is different; the nuchal dark blotch SERPENTS OF THAILAND 30. Lycodon laoensis Giinther. No. 336309, Chiang Mai, Chiang Mai, Thailand. Actual total length, 404 mm.; tail, 77 mm. 744 Tue UNIVERSITY SCIENCE BULLETIN Fic. 31. Lycodon laoensis var. No. 345296, Pattani, Pattani province. Actual total length, 335 mm.; tail, 69 mm. covers 17 transverse scalerows, and the eight succeeding ones cover 14, 14, 11, 9, 10, 9, 8, 6, transverse scalerows while the following ten body blotches cover either five or six; 14 blotches on tail; nowhere do spots or blotches reach ventrals; outer edges of sub- caudals may have some pigment. Variation: The known variation in ventrals is from 163 to 187; of subcaudals, 60 to 76. In Thailand specimens the ventral scales of the males vary from 165 to 187; the subcaudals, 60-73; females vary from 163 to 181, and 62 to 76. The Chiang Mai specimens have a slightly narrower occipital light band, and there appears to be a stronger constriction of the head on a level with the eyes; one specimen differs in having the prefrontal and frontal fused at a narrow isthmus. 745 SERPENTS OF THAILAND ‘apeu oq sAemje JOURS UNOS 9}eIND0R Ue 7eY} OS {ie} uo pue Apoq jo j1ed sJOL19}s0d oy} UL pauayIep oie sieq JYSI] Suynesedos ayy, ‘uoeutxoidde uy » + 1G cI Si GG Gl cI all IT As art aioe rer ET SOONG OA Kel GG. &G SG 9& 06 &G 9G IG Olea aE omnernr Apo seyozo[q Youpq sod a sod — soa sod sod == Se VaUrt angi eregony \SoSIv] [BSVU IOLIEJUy TT-O01 OI-OL 6-6 I-01 OTOL OI-OL O1-O1 O16 OL SOW a nin nes ae age EYOTA AE 6-6 OI-6 6-6 6-6 6-6 6-6 6-6 6-6 UES ONE pretense estes sperqeyeidng ze z z z z z z z GPR be oh ooo wa quINI04S0q I I I 1 I T T I To et ee quND0a1q ST-ZAT ST-LI ST-L1 STSLZI ST-Z1 ST-LI Sit 74 ST-LI ASSL tas ck re wien ys CENOO RON EMERY. z z z z z z z z 7 tees ee ee jeuy IZ 92 v9 HY) G9 G9 G9 v9 AEE verte lta ea Gen heats ee SER neOdnsS GLI AA Son SOT GLI Is] SLI 69T OO Tea Sa Eee or ees Seas SEEN 91 G tl & VI PF Gl FI 8&1 ial iat Al Watery Leet eo os bbe qysue] Petal Or I'8 SL Z 8 te) 8 8 Snape Pe | Wer ser ghoR Es agar og WEEE JOS SL LOT 06 08 69 82 GL 62 LL COI eH ee isa eles te. ok car Ree 9c TL S6P 68 OLE GEE 686 LEV GIV +0 STG end ee eee ering YySUd] BIOL, ip o ‘2 ‘2 Ws 4 4 8 0 Gono xog 646 6FS9E O&T9E 66SVE GE9EE 669EE Te9ee O&9ee SOEEE EACESOETUNL SISUBOD] uopooh'T uo eyep 9[BOS PUB S}JUDUTOINSCIINY 746 THe UNIVERSITY SCIENCE BULLETIN A specimen (No. 249) from Na Bon, Nakhon Si Thammarat was found precariously clinging to the almost-smooth surface of the trunk of a rubber tree about ten feet from the ground. Another was found under a small pile of dry grass. No. 249 contains a por- tion of a small Mabuya multifasciata. No. 36130, a male from Chiang Mai, contains a partly digested specimen of a Leiolopisma. Two Chiang Mai specimens (Nos. 33629, 33630) contain eggs, one four, the other three. No. 249 has an unusual parasitic worm embedded below the skin on one side. The parasite is approxi- mately 20 millimeters in length and from 1.5 to 2 millimeters in diameter. In the described specimen the loreal makes strong contact with the internasal while in most specimens it barely touches or does not touch the scale. There are fewer blotches on the body than in the other specimens. A specimen from the extreme southern border of the range, was taken in the city of Pattani in the province of the same name (No. 34529). It differs in having 36 dorsal bands with 22 on the tail. In this, the loreal is well separated from the internasal on one side (other side injured); the preocular is separated from the frontal on one side, touching only at a point on the other; the anterior tem- porals are superimposed and are of approximately the same size. The temporal formula is: 2 + 3-+4-+ 4. A second southern speci- men No, 3862 has been taken at Trang (Forest station ). Distribution: Found throughout Thailand. Outside of Thailand it is known from Malaya, Laos, Cochin China and Southern Viet Nam. Remarks: The specimen described was dislodged from under loose bark on a tree at an elevation of nearly 30 feet above the ground. Lycodon laoensis has developed certain variations which appear to be worthy of mention and _ sufficient southern material may permit subspecific designation for a southern form (figure given). The typical form was described by Gunther (1869) from Laos Mountains—two specimens “a male and a female, are perfectly alike.” There are 185 ventrals and 68 subcaudals in the male. “Abdomen and tail with a distinct angular ridge on each side in the male, indistinct in the female”; “preocular single, in contact with vertical [frontal] and with the third labial”; “two postoculars”; “a third specimen has been purchased and is said to be from Siam’; Boulenger (1893) adds the following details, “temporals 2 + 3”. 747 SERPENTS OF THAILAND “uOURT "I TRON » yepoy é CZ 69 OT CP é é é SUB, ZO8E CACHS 12 BAT LT 0g ne) &% 1z 1UB448q 6ZSFE 82S cg E91 g¢ ve) (Ee 9¢ uog ®N 6% OFZ TZ G/T OF ve) IZ CZ rey suey TS9EE EFS cg SLI SF 3} ral 9% rey Suvryy OSl9E 62S a) COT OF no) CT CZ ,OBL, Bud 10(] 6FS9E o4 OFS 9) eF9I 8¢ ne GT &% Tey suvryy 6Z9EE ES Z9 ISI SF ro) CT &% Tey SuvryyO Osage EZ £9 691 oe <) IT 1Z rey SuvryO ZEISS LEZ ¢9 CLT Ze vo) f6ll 0 JOT AN] SO8EE LEZ 17) 99] Ee ye) FI 61 UIRIG,, OWE 9Gz eL E81 é Oho geegeet ors é eaere “SYN SOV] CIN ECZ 89 CSI é Gara | hen case é , Soda, ‘Sy SOUT BINA LYG 89 6LI i Oe athe ee oa i SOTPI[BIO'T S.10q UN NY S[®1O T, nee S][BI}UO A ea xoG eae Gan: s1suaoD] UOPOIA'T Jo speaqUDA puR sayozoTq Apoq Surjsvsquod aIqv y, 748 THE UNIVERSITY SCIENCE BULLETIN Ventrals and subcaudals for the Siamese specimen are 183 and 73. The specimens in the preceding table have been arranged in a north-south series to contrast number of spots and ventral scales: One must regard the northern specimens as typical. In northern mainland Thailand the number of bands is from 19 to 26 on body; 11 to 15 on tail. In the southern (peninsular) part of Thailand, two specimens have 25 and 36 on body; 21 and 23 on tail. Total number of bands in northern Thai forms, 33 to 40; southern Thai forms, 46 to 58. Mr. Boo-Liat of Kuala Lumpur has reported (Malayan Nature Journal vol. 12, 1958, pp. 183-184, pl. 12) two specimens from Kedah one of the northernmost Malayan states. His figure does not show as numerous caudal spots as occurs in the southern Thai specimens. Lycodon capucinus Boie (Fig. 32) Lycodon capucinus Boie, Isis, 1827, p. 551, based on Russell’s Indian Serpents, vol. 2, pl. 39, (type-locality, Trivandrum India); Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, p. 1136 (Rat Buri province, Thai- land). Lycodon aulicus capucinus Smith, The fauna of British India, Ceylon, Burma, including the whole of the Indo-Chinese Subregion; Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, pp. 265-266; Boulenger, A vertebrate fauna of the Malay Peninsula. . . Reptilia and Batrachia, 1912, pp. 131-132 (part.). [Many references to Lycodon aulicus belong here but it is not always possible to determine which of the two forms is indicated. ] Lycodon aulicus Smith, Journ. Nat, Hist. Soc. Siam, vol. 1, no. 1, p. 16-17 (Lop Buri, Chiang Mai, Chon Buri). The two forms presumed subspecies of aulicus occur together throughout much of their ranges. It seems necessary to regard them as species. Diagnosis: Purplish to purplish brown with a reticulum of white or fawn, color confined to edges of scales and interstitial skin; snout and head rather flattened; a white band across occiput and nape, sometimes partly or completely divided mesially, the scales with lavender or purple spots; labials white with black dots; ven- trals kneeled, ends turned up on ventrolateral area; loreal large, forming a suture with internasal; nine supralabials, three bordering orbit. Description of species: (From No, 34948. Phu Kading, Loei province, Thailand): Snout flattened, projecting beyond mouth, somewhat excavated below, with a transverse blunt edge anteriorly; rostral higher than wide, bent back over tip of snout, well visible SERPENTS OF THAILAND 749 Fic. 32. Lycodon capucinus Boie. From Thailand. Total length, 528 mm.; tail, 93 mm. above; internasals much smaller than prefrontals, slightly angular in front, forming a suture with loreal; prefrontals distinctly longer than wide, separated from supraocular by a minute distance; frontal slightly shorter than its distance from tip of snout, distinctly shorter than parietals; anterior nasal larger than posterior; loreal more than twice as long as high; large preocular touching frontal; two post- 750 THE UNIVERSITY SCIENCE BULLETIN oculars, each in contact with a temporal; two anterior temporals, 1 1 the formula on right side —-+ 2+ 3, on left, —+ 2-44; four 1 1 temporals border parietals, the last separated behind parietals from its fellow by three scales; nine supralabials, third, fourth, and fifth enter eye (the third only narrowly); ten infralabials, first five touching first pair of chinshields, which are much larger than sec- ond pair. Scales smooth, some with a minute apical pit (that can scarcely be discerned except in a transparent shed scale). Scale-formula 17-17-15. Ventrals, 205; subcaudals, 71; anals, 2; lateral keels on ventrals moderately distinct. Color in life: Head dark purplish back to level of parietals; parietals light brownish lavender; occipital light band not divided mesially, each scale with a purplish spot; this followed by a brown band; labials all yellowish with a large purplish brown spot; body lilac to lavender-brown with a reticulum of whitish or fawn (this chiefly on scale edges and interstitial skin); tail nearly uniform brown; dirty white on venter. Measurements and data on Lycodon capucinus NUMBER 34948 2 35954 33633 No No. Total length......... 521 436 385 528 Bann asec yanet A ae 113 73 47 93 Head, width......... 10.2 9 9 10 Head, length......... 18 14.6 16 15.9 Vienthallssevsr et a 205 195 188 203 Subcaudals........... 71 61 67 65 Preoculars........... 1 1 1 1 Postoculars.......... 2 2 2 2 Supralabials.......... 9-9 9-9 9-9 9-9 Infralabials.......... 10-10 10-10 10-10 10-10 Anterior nasal largest. . yes yes yes yes Labials enter eye...... 3, 4, 5 3, 4, 6 3, 4, 5 3, 4, 5 (4, 5) | No. 34948 from Loei; 35954, 33633, Chiang Mai. Variation: The specimens available agree in most characters. However one specimen has 19 rows of scales about neck, seventeen being the normal number. Ventrals vary in Thai specimens from 188 to 205, the subcaudals from 61-71, the higher numbers of ven- trals and subcaudals being found on females. Distribution: The species is known in Thailand from Chon Buri, Lop Buri, Rat Buri and Chiang Mai. Outside of Thailand the SERPENTS OF THAILAND 751 species is known from India throughout southeastern Asia, Malay Peninsula, the Indo-Australian Archipelago to the Philippines and Celebes. Remarks: The species is very domestic, being found in human habitations, especially those built with bamboo and palm leaves. It is presumed that the domestic geckoes attract these snakes since lizards form the major food of the species. Lycodon fasciatus (Anderson) (Fig. 33) Ophites fasciatus Anderson, anatomical and zoological researches and zoological results of the Yunnan Expeditions. Reptilia collected on the two expeditions to western Yunnan, 1879, p. 827, pl. 78, fig. 1; (type-locality Ponsee [Pangsi] Yunnan ); Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 614. Lycodon fasciatus: Boulenger, The Fauna of British India including Ceylon and Burma, 1890, pp. 295-296; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, p. 358; Wall and Evans, Journ. Bombay Nat. Hist. Soc., vol. 13, 1900, p. 372; Evans, ibid., vol. 30, 1925, p. 812; ibid., vol. 31, 1926, p. 562; Schmidt, Bull. Amer. Mus. Nat. Hist., vol. 54, 1927, p. 523; Pope, Reptiles of China, 1935, pp. 188-191; Bourret, Le ser- pents de l’Indochine, 1936, p. 155; M. Smith, The fauna of British India including Ceylon and Burma, Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, pp. 266-267. Diagnosis: A snake reaching nearly a meter in length. Black bands, wider anteriorly, separated by narrow yellowish bands that are sparsely pigmented with brown or black. Head black-brown above, lighter on temporal and occipital regions (white in young). An elongate loreal entering orbit with a small preocular above it, separating prefrontal from eye; scalerows, 17-17-15, the median five to seven or more rows keeled; ventrals feebly angulate. Description of species: (Specimen from Doi Suthep, Chiang Mai, Thailand ): Head somewhat flattened, snout projecting a little be- yond mouth; temporal region somewhat swollen; rostral distinctly broader than high, portion visible above very narrow; internasals about a third of area of prefrontals, about as wide as long; pre- frontals large, their common suture nearly double that between internasals; frontal a little shorter than its distance from tip of snout, very much shorter than parietals, four-fifths as wide as long; nostril between two nasals posterior much the larger with a de- pression behind nostril; loreal elongate more than two-and-one-half times as long as high, entering eye; a preocular separating pre- frontal from eye; supraocular widened posteriorly, about half width of frontal; temporals, 2 + 2-+ 2 the upper anterior touching two postoculars; supralabials eight, third, fourth, and fifth entering orbit; nine infralabials, five touching first pair of chinshields which THe UNIVERSITY SCIENCE BULLETIN Fic. 33. Lycodon fasciatus (Anderson). Specimen from Doi Suthep, Chiang Mai, Thailand. Actual total length, 830 mm.; tail, 169 mm. SERPENTS OF THAILAND 753 are a little larger than third pair; second pair very small. Scale- formula: 23 (occiput), 17, 17, 15, with small apical pits; anteriorly five median scalerows lightly keeled, while on posterior half of body seven to nine rows more distinctly keeled; opposite vent there are 13 rows strongly keeled. Ventrals, 209; subcaudals, 85; anal, single. Color in life: Banded purplish black and cream yellow, the yellow bands two to three scale lengths in width, the edges very irregular on venter, widening so as to cover from three to five ventrals on anterior half of body, one or two on posterior part of body; scales in the yellow-cream bands pigmented more or less, often form- ing small spots or indefinite blotches, more strongly indicated on latter two thirds of body and tail; black bands vary in width, the nuchal band covering 18 transverse rows, and the succeeding eleven ones cover 13, 11, 10, 8, 7, 6, 6, 7, 7, 6, 5 respectively on the mid- line; the following 16 body bands cover between 3 and 5 trans- verse rows. The 13 bands on tail usually cover between three and five. Head brownish, anterior part darker than posterior; supra- labials each with a dirty cream center; chin with brownish pig- ment, posterior infralabials cream; first seven ventrals cream. _ Measurements in mm.: Total length, 830; tail, 169; snout to vent, 661; width of head, 18; length of head, 19. Variation: The head is distinctly wider than the neck. The ven- tral scales are angulate laterally (but not strongly so) and the body is somewhat compressed. The tail tapers to a fine point. A younger specimen examined has thirty dark bands on body, eighteen on the tail. On the head a dark brown mark covers the snout and the interorbital area, terminating on the anterior fourth of the parietals. The occipital and temporal regions are light brown, bordered behind by a narrow white line. A few of the bands on the anterior part of body are broken on the venter. The squamation is as follows: supralabials, 8; infralabials, 9; there are 3-2 postoculars, the temporals, 2+2+3, 2+3-+3. There are three pairs of chinshields, the first pair bordered by four infralabials, the second, smallest pair bordered by two, the third pair touching only one infralabial. There are 206 ventrals, 80 sub- caudals, and the anal is single. In the described specimen the left anterior chinshield is divided, but the middle chinshield on the right side is fused to the posterior. The color of the light bands is gray-white, the ventral coloration being somewhat cream. 754 THE UNIVERSITY SCIENCE BULLETIN The known variation in the ventrals is from 197 to 220; in the subcaudals, 69 to 94. The number of body bands varies between 28 and 49. In the very young, the back of the head is white. Distribution: In Thailand the species has been reported from Tawkawbee, 9 mi. S. of Um Pang (lat. 16 N. long. 98° 75’ E.). Both specimens here reported are from Doi Suthep, Chiang Mai. The species extends from the eastern Himalayas, to Tibet, Burma, upper Laos, West China, and Viet Nam. Remarks: The species is a bush snake, usually found at eleva- tions of from 1000 to 2300 meters. The largest specimen known to me reaches a length of 934 mm. Genus Dinodon Duméril and Bibron Dinodon Duméril and Bibron, Mem. Acad. Sci. Paris, vol. 23, 1853, p. 463; Erpétologie générale . . . vol. 7, 1854, p. 447, (type of genus, can- cellatum = rufozonatum ). Eumesodon Cope, Proc. Acad. Nat. Sci. Philadelphia, vol. 12, 1860, p. 262, (type of genus, semicarinatus ). Lepidocephalus Hallowell, Proc. Acad. Nat. Sci. Philadelphia, vol. 12, 1860, p. 498, (not of Bleeker), (type of genus, semicarinatus ). Adiastema Werner, Sitz. Ber. Akad. Wiss. Wien, vol. 134, p. 54, (type of genus, cervinum ). Lycodon Boulenger, The fauna of British India . . . Reptilia and Batra- chia, 1890, p. 291 (part.). Diagnosis: Head rather broad, somewhat flattened, scarcely dis- tinct from neck; eye with a vertical pupil; scale-formula: 17, 17(19), 15; scales smooth or with indistinct keels; ventrals with or without lateral keels; scales with apical pits. Maxillary bones extending farther forward than palatine, bent inward, with five to seven anterior teeth increasing in size pos- teriorly, sometimes separated from posterior teeth by a diastema; posterior teeth of first group fanglike, of second group posterior- most only slightly larger than others. Of the nine recognized species only a single one is known to reach Thailand. It enters from the north and has been traced as far south as Chiang Mai. It is difficult to differentiate between Lycodon and Dinodon since the differences are small. The shape of the maxillary bone and the dental characters are perhaps the most important. The wisdom of retaining Dinodon as a separate genus has been questioned. Dinodon septentrionalis septentrionalis (Gunther ) Ophites septentrionalis Giinther, Proc. Zool. Soc. London, 1875, p. 233 (type- locality “there is no doubt that he [Dr. Jerdon] obtained it during his last journey through the northern parts of India”). SERPENTS OF THAILAND 755 Lycodon septentrionalis Boulenger, The fauna of British India . . . Rep- tilia and Batrachia, 1890, p. 295 (states the length of type 83 inches (evi- dently a misprint ) should be 33 inches). Dinodon septentrionalis Boulenger, Catalogue of the snakes in the British Museum, vol. 1, 1893, p. 363, (part.), and vol. 3, addenda, 1896, p. 619 (Cobapo, Karin Hills); Ann. Mus. Civ. Genova, ser. 2, vol. 13, 1893, p. 324; Journ. Bombay Nat. Hist. Soc., vol. 16, 1905, p. 235; Wall, Journ. Bombay Nat. Hist. Soc., vol. 18, 1908, p. 778; ibid., vol. 29, 1923, p. 615; Rec. Ind. Mus., 1909, p. 146; Angel, Bull. Mus. Nat. Hist. Paris, 1929, p. 79; Bourret, Serpents de lIndo-Chine, 1936, p. 162; M. Smith, The fauna of British India, Ceylon and Burma. including the whole of the Indo-Chinese Sub- region; Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 270-271, fig. 91. Diagnosis: Black with 25 to 35 narrow white or yel!owish bars, becoming wider on sides; venter whitish; tail black below; 207-212 ventrals with lateral keels; 81-92 caudals; anal, single. Description of species (composite sources): Rostral narrowly visible above, slightly angulate on median line; internasals some- what narrowed in front; prefrontals about two and a half times size of internasals; suture between prefrontals and frontals nearly a straight line; frontal as broad as long; supraoculars slightly more than half width of frontal; parieta's elongate, but length less than distance between posterior part of frontal and tip of snout. Nasal divided, posterior portion higher and larger than anterior; loreal small, narrow, separated from internasal and eye; two small post- oculars; two anterior temporals followed by two posterior; lower anterior and upper posterior temporal much the largest. Eight supralabials, third, fourth, and fifth bordering orbit; seven or eight infralab‘als. Scales smooth, the formula: 17, 17, 15. (One presumed sub- svecies may have from five to seven median rows keeled. ) Color: Purplish black above and on sides, with 25-35 narrow white transverse bars on body which may expand laterally; on forepart of body bars about twice as far apart from one another as on hinder part; venter white, sometimes spotted or barred with black, these markings being a continuation of the dark color on back; tail heavily marked with black below; posterior part of head white in young, usually black in adults. Measurements in mm.: Total length, 1,180; tail, 190. Distribution: The species has been found only in the northern part of Thailand in the province of Chiang Mai. Outside of Thailand it occupies territory in the Eastern Hima- layas, Assam, Burma, and Laos. Remarks: The type of Dinodon septentrionalis (Gunther ) is said to be a snake, 83 in. (1,962 mm.) in length according to Ginther.* ® Boulengcr, |. c. 1893, believes this is intended for 33 inches. 756 THe UNIVERSITY SCIENCE BULLETIN There are 214 ventrals and 83 subcaudals. It is black in color, the trunk with thirty narrow white rings only two scales wide, the first at some distance behind the head. It is white on venter, the sub- caudals marbled with black. Ventrals @ 214; subcaudals, 83. Genus Oligodon (Boie in Boie) Oligodon Boie in Boie, Isis, 1827, p. 519 (type of genus Oligodon bitorquatus). Simotes (not of Fischer 1817) Duméril, Bibron and Duméril, Erpétologie général . . . vol. 7, 1854, p. 624 (type of genus, russelli). Holarchus Cope, Proc. Amer. Phil. Soc., vol. 23, 1886, p. 488 (type of genus not designated ). Dicraulax, Cope, Amer. Nat., 1893, vol. 27, p. 480 (type of genus Holarchus trinotatus = Simotes purpurascens. Malcolm Smith writes (1943, The fauna of British India, pp. 196- 197): “Wall quite rightly (1923) has united Holarchus with Oli- godon, the latter being only a degenerate group of the former. The passage from one to the other is gradual and no dividing line can be drawn. Degeneration has led to reduction in the number, but not always in the size, of the maxillary teeth; reduction in the number and size of the palatine teeth, but in no species are they entirely lost.” Diagnosis: Rather short medium-sized snakes rarely reaching a meter in length, while many of the species are less than half a meter. Maxillary teeth 6-16; scalerows, 13-23. Key To THAI SPECIES AND SUBSPECIES OF Oligodon iw) 1° Sealesswith paired*apical pits... 4.222. ss 2 1 Scalestwithout .apical’plts: 2.20. a. .6 2 2he eee es ee . Seales in 17 rows at mid-body; spots on venter present or absent (two SUDSPECIES) cares cath bis see Suse rye ee eye ni eet ee cinereus Seales in 15 rows at mid-body; spots absent on venter........ tnornatus 3. Scales in’2)- rows at: mid-body<. 2.22.2 ss he eee + Scales in 19 rows or less at mid-body.............. 0.00000 cece e eeu aee 6 4. Body with two median darker longitudinal stripes separated by a lighter median dorsal stripe; a lateral stripe to vent level; ten more or less divided dorsal dark blotches; tail long, one fifth to one sixth of total length; venter immaculate; ventral-subeaudal count high (to 22914)... .dorsolateralis Body without median or lateral darker longitudinal stripes; markings TPANISVICTSO i ce ove oi sk ee Ee seta et etic TE TIPS eTe ea 5 5. Series of 13-18 transverse dark spots on back and tail or spots with white or light centers, separated by ground color and three narrow irregular dark transverse lines, the dark color confined to edges of scales; venter without black quadrangular marks: ~.. 0.3.00... 52 ee cyclurus smuthi No distinct spots but numerous transverse irregular lines across dorsum, the dark color confined to edges of scales........... cyclurus superfluens 6-Seales:.in: 19) rowstv6s 3a we ete cate eae eee Scales in 17 to 15 rows; more or less than half a meter in length........ 8 OO i) “I SERPENTS OF THAILAND 757 7. Small slender snakes with four dark longitudinal lines on body, the two median separated by a lighter line; a narrow lateral line; no spots on dor- Salvoncau dal tancaee sas sce ay een grenmises ty yiease ieee cos quadrilineatus Larger snakes reaching about 90 em. in length without longitudinal dorsal stripes but with transverse dark spots and narrow dark lines (rarely 21 S Call CO Wis) Berea oe sete rns ee Reon eee pe nee eee p. purpurascens 8. More than half a meter in length (to 760 mm.); scalerows 17 or 15 near mid-body; no longitudinal lines. Maxillary teeth twelve, steadily increas- ing in size posteriorly. Purplish-brown above with ill-defined and broken cross-bands, every alternate one enlarged. Belly red or red barred with black (in life); head markings like cyclurus; hemipenis like cinereus. joynsont Less than half a meter in length; scales in 17 rows; longitudinal lines on hyo Glivaeitoeo na menaan eee clerk eeataenidra ni Wenn pata Acree ao ant hun cmaie mel angen a un 9 9. Maxillary teeth 14-16; ventrals, 146-169; subcaudals, 30-47. Four dark brown or blackish longitudinal lines, two median (wide) separated by a somewhat irregular narrow light stripe (width of one scalerow); series of quadrangular spots on ends of ventrals; ventrals coral-red in life, whitish in preservative. A black spot at base of tail and one at tip (rarely ab- (SETI) Shae fees RC ls a a EEE en ie nar eT taentatus Maxillary teeth 13-14; ventrals 135-160, subcaudals 32-44; 14 or 15 trans- verse, rather narrow dark dorsal spots, light-edged on body and tail (often broken in two or four separate dots) separated by three more or less dis- tinct transverse marks, the black confined to edges of scales; a dark-edged median light longitudinal line; venter with quadrangular spots. . . barron Oligodon cinereus (Ginther ) Simotes cinereus Giinther, The reptiles of British India, 1864, p. 215 (type- locality, “Gamboja” = Cambodia). This species, extending from the Indo-Chinese region to India and southern China, seemingly has broken up into several sub- species. At least two forms occur in Thailand, one in the northern part of Chiang Mai province, and one in the southern part of that province and in Loei province. These are Oligodon c. swinhonis and Oligodon c. multifasciatus. They may be distinguished by the following key: Key ro THAI SUBSPECIES OF Cinereus 1. A well-defined narrow chevron on neck, the posterior arms extending from three to four scale-lengths posteriorly; a narrow anterior process from chevron terminates on frontal (sometimes forming two small spots); diag- onal black stripe above mouth angle; and a band across snout through eyes; venter yellowish white with paired series of quadrangular dark spots on outer sides of ventrals, those bearing spots separated by one or two lacking spots; 17 scalerows at middle; ventrals 190-197; subecaudals 32- a ieee ilst © aie eel ope asols ay oe eoe Ulea h rie Ue ee Oe TNE ERR cinereus multifasciatus . Dark chevron on neck dim the centers of scales involved light gray; band across snout almost or completely obsolete or a dim dark spot below eye; diagonal stripe above mouth-angle usually absent; entire chin and venter immaculate, the upturned edges of posterior ventrals with some pigment. Part of the transverse marks on body intensified somewhat, separated by three dim transverse marks, the middle one may be a little darker than OLIVER GW Oe aie dee ike othe a ator RUD RNS a | NH Sah ety ehcp ci acs at cinereus swinhonis Ww 758 THe UNIVERSITY SCIENCE BULLETIN Oligodon cinereus swinhonis Gimnther (Fig. 34) Oligodon swinhonis Ginther, The reptiles of British India, 1864, p. 215, pl. 20, fig. E (type-locality, Amoy, China). Simotes violaceus: Boulenger, Catalogue of the snakes in the British Museum (Natural History), vol. 2, 1894, pp. 222-223 (part.). Oligodon cinereus: M. Smith, The fauna of British India, Ceylon and Burm, : Reptilia and Amphibia, vol. 3, Serpentes. Dec. 1948, pp. 215-217 (part.). Diagnosis: See key. Description of subspecies (from No. 3464 ¢ , Chiang Mai prov- ince): Rostral higher than wide, the part visible above triangular, its length greater than its distance from frontal, the scale thickened, its edges elevated on sides; internasals triangular, much narrowed posteriorly, their common suture about half as long as that between prefrontals; latter scales twice as wide as long; frontal shield- shaped about one fifth longer than wide, longer than its distance from tip of snout, shorter than parietals, less than twice as wide as supraocular; parietals sharply truncate behind, somewhat emarginate laterally; nasal divided; loreal longer than high; a single preocular, and on left side, a subocular, none on the right side; two postoculars; temporals, 2+2+3 (2+ 3-+3); supralabials seven, (the fourth and fifth presumably fused) and only a single scale bordering orbit; infralabials nine, four touching first chin- shields which are double size of second pair. Scale-formula: 23 (occiput), 17, 17, 15; scales smooth with distinct apical pits. Ventrals, 196; subcaudals, 42; anal, single. Tail definitely rounded above, flat below. Color: Above generally brownish fawn to gray (dull salmon- pink in life); anteriorly outer scalerow nearly salmon-pink prac- tically without pigment, but more posteriorly covered with gray pigment. A dark chevron on nape the posterior arms of which equal in length to three or four scales; and with a forward process reaching onto parietal suture. A faint dark mark below eye but no trace of the band through eye across snout; no diagonal marks above angle of mouth. Edges of many scales on dorsum dark forming irregular trans- verse marks reaching on sides to outer scalerow. One series of about 21 transverse marks somewhat intensified, separated by three intervening bands, the median one darkest. Median scalerow on body vaguely lighter than others and on tail this line, with in- definite edges, is more distinct. Venter immaculate yellowish white SERPENTS OF THAILAND 759 Fic. 34. Oligodon cinereus swinhonis Giinther. No. 3464, 12 km. norih of Chiang Dao, Chiang Mai province, northern Thailand. Actual total length, 632 mm.; tail, 84 mm. to ivory lacking ventral spots of black; outermost edges of ventrals posteriorly have some pigment. Measurements in mm.: Total length, 632; tail, 84, width of head, 14; length of head, 18.2. Variation: The type, a'so a female, has 174 ventrals, and 34 sub- caudals. The described specimen has considerably more than this number—196, and 42. In Thailand this is, so far as I am aware, the only specimen that has been taken. It is presumed that it is a northern form, occuring in southern China (Amoy and westward). 760 THe UNIVERSITY SCIENCE BULLETIN Oligodon cinereus multifasciatus Jan (Fig. 35) Oligodon multifasciatus Jan in Jan and Sordelli Iconographie générale des ophidiens, vol. 1, livr. 12, pl. 5, fig. 2, (type-locality “Sultanepore” ). Simotes violaceus: Boulenger, Catalogue of the snakes in the British Museum, vol. 2, pp. 222-223 (part.). Oligodon cinereus: M. Smith, The fauna of British India, Ceylon and Burma i vol. 3, Serpentes, Dec. 1943, pp. 215-217 (part.). Diagnosis: See key. Description of subspecies (from EHT-HMS No. 3623, Chiang Mai province, Thailand): Part of rostral visible above forming a triangle, the scale thickened, slightly elevated when seen from sides; internasals much wider than long, somewhat pointed mesially; prefrontals with a median suture less than twice length of inter- nasal suture; frontal shield-shaped, about as long as wide, shorter than its distance from tip of snout, shorter than parietals; posterior border of parietals forming a median angle. Nostril between two nasals, anterior one the larger; loreal longer than high touching two labials; a large upper preocular, and a small lower (subocu- 9 lar); two postoculars; temporals, ae + 2, somewhat irregular; supra- 1 labials, 7-7 (due to apparent fusion of second and third on one side, and of fourth and fifth on other); infrabials, 8-8, four touch- ing anterior chinshields which are more than twice as large as second pair. Scale-formula: 21 (about occiput), 17, 17, 15 (just preceding vent). Ventra!s 196; subcaudals (paired), 38; anal, single. Color: Ground color brownish gray with black edges of many scales tending to form irregular broken transverse marks separated by one scalerow; edges of the intervening row usually cream or ivory, the color normally concealed and extending to interstitial skin. A narrow chevron on nape, its posterior arms equal in length to four or five scale-lengths; anterior median projection broken, with a round spot on parietals and one on frontal; a dim dark diagonal line above angle of mouth; a line beginning on labials passes through eye and across forehead anterior to frontal. Venter distinctly yellowish ivory with a double row of somewhat quad- rangular black spots on ventrals, paired or alternating, usually separated by an unspotted ventral; chin unspotted; a few subcaudal spots; tail rounded above, flat below with an indistinct median lighter supracaudal line. Measurements in mm.: Total length, 683; tail, 77; width of head, 13.2; length of head, 20. Fic. 35. Oligodon cinereus multifasciatus Jan. No. 3623, Chiang Dow (12 km. north), Chiang Mai province. Actual total length, 683 mm.; tail, 77 mm. 762 THe UNIVERSITY SCIENCE BULLETIN Variation: The following table shows variation in three Thai specimens: Scale-data and measurements of Oligodon cinereus multifasciatus r . Ven- Sub- Supra- | Total Tail - dp aa) NUMBER Erovince trals | caudals | labials | length | length SOOO Tooele 4 OR ie 190 32 8-7* 600 71 SOLS eee Chiang Mai.... 196 38 7 683 77 PA Nie Sesh oh Chiang Mai... ‘| 197 | 42 7 90 Zl | * The 2nd and 3rd scales are fused on one side; there are two scales bordering orbit; the tail of No. 31699 is immaculate below. Remarks: The form is known in Thailand from southern Chiang Mai and Loei provinces. The type-locality is “Sultanepore,” a locality unknown to me. Oligodon inornatus (Boulenger ) Simotes inornatus Boulenger, Journ. Nat. Hist. Soc. Siam, vol. 1, pt. 1, 1914, p. 68 (type-locality, Nong Kai Ploi, E. of Sriracha, southeastern Thailand); M. Smith, ibid., pp. 68-69. Oligodon cinereus M. Smith, The fauna of British India . . . vol. 3, Serpentes, 1943, pp. 215-217 (part.). Diagnosis: Uniform pale brown above, yellowish white beneath scales in 15 rows about body. Description of species (from Boulenger’s type-description ): Nasal divided; portion of rostral seen from above as long as its distance from frontal; suture between internasals longer than that between prefrontals; frontal equal to its distance from tip of snout, and as long as parietals; two postoculars, no suboculars; temporals, 1+ 2; eight supralabials, fourth and fifth bordering orbit; four infralabials bordering anterior chinshields which are nearly twice as long as posterior. Scales in 15 rows; ventrals, 171; anal undivided; subcaudals, 42. Color: Uniform pale brown above, yellowish white beneath. Measurements in mm.: Total length, 580; tail, 90. Remarks: Published with Boulenger’s paper (loc. cit.) was a description of a specimen of Oligodon by Dr. Malcolm Smith, which he regarded as a second specimen of Boulenger’s inornatus. Smith (1943) abandoned the species placing inornatus as a synonym of Oligodon cinereus (Gimther ). SERPENTS OF THAILAND 763 However there is a very strong possibility that inornatus is worthy recognition; and the specimen referred to it originally by Smith may require a reassignment. Smith’s statements regarding this form follow: “This snake was collected and kindly forwarded to me by Mr. P. A. R. Barron of the Borneo company, to whom I am also indebted for a second specimen, which differs from the type description in the following particulars. Position of rostral seen from above, shorter than its distance from the frontal; suture between the internasals equal to suture between the prefrontals. Loreal as long as deep. Ventrals 174. Subcaudals 40. Color. Above, dull, salmon red (much brighter I am informed in life) and with very faintly marked dark greyish narrow cross-bars. Below, with indistinct, small squarish spots, placed laterally, in the posterior half of the body and tail. This specimen has been in my possession for some months, and the handsome red coloring has gradually faded until now it is nearly of the brown hue described by Mr. Boulenger. Mr. Barron who saw both specimens when fresh informs me, however, that the first one never showed any red coloration.” The type locality of inornatus is Nong Kai Ploi, East of Sriracha, Chon Buri. Boulenger regarded the species as being related to violaceus. M. Smith (1943) remarks that, “Specimens from the extreme south-eastern corner of Siam (south of Retriu) and east- wards to the adjacent territory of Cambodia, have only 15 scale rows at mid-body.” Oligodon dorsolateralis Wall Simotes albocinctus dorsolateralis Wall, Journ. Bombay Nat. Hist. Soc., vol. 19, p. 898 (type-locality Jalpaiguri District, no type designated). Oligodon cyclurus: M. Smith, The fauna of British India, Ceylon and Burma, . . . Reptilia and Amphibia, vol. 3, Serpentes. Dec. 1943, pp. 202-206 (part. var. IV.) Diagnosis: Scales in 21 rows, with four longitudinal stripes, the two median extending to tail, the two lateral to level of vent; ten more or less divided dorsal blotches, four on tail. Tail long, be- tween one fifth and one sixth of body length; venter immaculate, occasionally spotted (in Burmese specimens ). Description of species (from No. 77 g ): Rostral about as high as wide, the triangular part visible above, distinctly shorter than its distance from frontal, thickened as seen from sides; internasals rather angular on anterior edge, much wider than long, somewhat 764 THe UNIVERSITY SCIENCE BULLETIN obliquely placed, prefrontals nearly twice as wide as long, their common suture equal to that between internasals; frontal slightly longer than its distance from tip of snout, as long as parietals; latter nearly as broad as long; nasal divided, anterior part largest; loreal quadrangular, higher than long; two preoculars, or one preocular and one subocular, latter lying below level of eye be- tween third and fourth labials; two postoculars, both touching first temporal; temporal formula, 1 + 2 -+ 3, somewhat irregular (one might regard the arrangement as two anterior temporals the lower largest and separated from both postoculars by the upper); eight supralabials, fourth and fifth bordering orbit; nine infralabials, four touching anterior chinshields which are somewhat more than twice size of second pair. Scale-formula: 24 (occiput), 19, 21, 17; ventrals, 176%; anal, single; subcaudals, 53. Body somewhat angular ventrally; tail distinctly triangular in cross-section and somewhat stiffened, ending in a semirigid point. Color in life: Above light brown with a pair of median stripes, each covering two scalerows and part of a third, posteriorly separated by a light median stripe covering all of one scalerow; posteriorly stripes extending to tip of tail, anteriorly joining a chevron-shaped mark that terminates at middle of frontal; two oblique temporal marks cross angle of mouth, passing slightly be- yond; a wide dark mark on snout reaching from back part of frontal to rostral, laterally crossing eye to upper part of fifth labial; a lateral stripe on edges of fourth and fifth scalerows, reaching to level of vent; ten dorsal blotches mesially divided and more or less con- fluent with small lateral spots on lateral stripe; below on chin and venter immaculate ivory; ventrals upturned laterally, the upper edges with some grayish pigment similar to adjoining scalerows. Tail with four dorsal spots. Measurements in mm.; Total length, 607; tail, 105; width of head, 13; length of head, 20. Variation: The specimen described is from the city of Chiang Mai. I first saw the specimen from a hospital window in the yard and captured it. The species originally described by Wall, was placed as a sub- species of albocinctus; Malcolm Smith regards it as a form of cyclura. As another subspecies of cyclura occurs here it seems best to regard this as a distinct species. SERPENTS OF THAILAND 765 Distribution: This form is known only from this locality in Thailand. Elsewhere it occurs in Burma and Assam or near the Assam-Bengal boundary. Oligodon cyclurus (Cantor ) Coronella cyclura Cantor, Proc. Zool. Soc. London, 1839, p. 50, (no type- locality known, the type lost. A colored drawing signed by Cantor exists. ) Simotes cyclurus: Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 311; Catalogue of the snakes in the British Museum, vol. 2, 1894, pp. 219-220; Ann. Mus. Civ. Genova, ser. 2, vol. 13, 1893, pp. 324-325: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, 1914, p. 97, fig.; Wall, Journ. Bombay Nat. Hist. Soc., vol. 18, 1908, p. 780. Holarchus purpurascens: Cochran, Proc. U. S. Nat. Mus., vol. 77, 1930, pp. 27-28 (Bangkok). Holarchus cyclurus: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 14, 1920, p. 96. Oligodon purpurascens: (non Schlegel) Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 631; ibid., vol. 30, 1925, p. 815; ibid., vol. 31, 1926, p. 563; Rec. Ind. Mus., vol. 35, 1923, p. 328; ibid., vol. 25, 1923, p. 328; M. Smith, Bull. Raffles Mus., no. 3, 1930, pp. 53-55. Simotes fasciolatus Giinther, The reptiles of British India, 1864, p. 218, pl. 20, fig. B (type-locality “Petchabun” southeast Siam (possibly Phet Buri (Petchaburi) SW Thailand at the upper end of the peninsula ). Simotes smithi Werner, Sitz. Ber. Akad. Wiss. Wien, vol. 134, 1925, p. 58 (Siam); M. Smith, Ann. Mag. Nat. Hist., ser. 10, vol. 1, 1928, p. 497. Coronella violacea Cantcr, Proc. Zool. Soc. London. 1839, p. 50, (type-locality Rangpur, Bengal). Oligodon cyclurus: Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Dec. 1943, pp. 202-205, figs. 62-63 (part.); Taylor and Elbel Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, pp. 1136-1137 (Chaiyaphum, Loei and Nakhon Phanom provinces ). Diagnosis: A rather large species of the genus reaching a length of 940 mm. variable, producing a variety of forms that should be treated as subspecies; scale formula, 21, 21, 17. Oligodon cyclurus smithi (Werner ) (Fig. 36) Simotes smithi Werner, Sitz. Ber. Akad. Wiss. Wien, vol. 134, 1925, p. 58, (type- locality, “Don-Rek,” Siam, presumably the Dong Rek mountains between Thailand and northern Cambodia. ) Oligodon cyclurus: M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 202-204, fig. 62; Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, pp. 1136-1138, fig. 25. Tee purpurascens: Cochran, Proc, U. S. Nat. Mus., vol. 77, (2) 1930, Di oie Diagnosis: Brownish or yellowish brown with 15-17 black or brown spots edged with darker color, each spot separated by three narrow irregular black transverse marks, the color broken and confined to edges of scales. A strong chevron on neck extending to frontal, its posterior branches reaching about to level of 10th 766 Tue UNIVERSITY SCIENCE BULLETIN Fic. 36. Oligodon cyclurus smithi (Wemer). EHT-HMS No. 317014, Kho (Mt.) Kanbuang, Nakae district, Nakhon Phanom province, Thailand. Actual total length, 579 mm. ventral; a mark across snout and a stripe from occiput across jaw- angle; venter immaculate. Description of species (from No. 33569, Ang Hin, Chon Buri): Rostral considerably wider than high, thickened and bent back SERPENTS OF THAILAND 767 over tip of snout, portion visible above triangular, its length dis- tinctly greater than its distance from frontal; internasals diagonally placed, their mutual suture about one fourth of their diagonal length (width); prefrontals likewise narrowed greatly on mid-line their mutual suture about a fourth of the width of scale; frontal about one sixth longer than wide, its length equal to its distance from end of snout and equally as long as parietals; parietals as wide as long; nasal divided, anterior part the larger; loreal about as high as long; two preoculars, lower small, between third labial and eye; two postoculars; temporals, 2+ 2+ 3 on right side (irregular on left side); supralabials, eight, fourth and fifth border orbit, last two largest; greatest diameter of eye much less than its distance from lip, less than its distance to nostril. Mental tiny; nine infra- labials; first pair of chinshields twice as large as second, bordered by four infralabials. Scale-formula: 24 (occiput), 17, 17, 15; ventrals, 162; subcaudals, 49: anal, undivided. Color: Above gray-brown; body with 13 or 14 somewhat faded spots on body; spots on tail obsolete; between spots three narrow dark transverse bars; black color not reaching below second scale- row; venter immaculate; nape with a large chevron-shaped mark terminating on frontal, its posterior branches terminating at level of eighth ventral; a diagonal bar from parietal across jaw-angle; a band across snout widened mesially passing back and down through eye and across fourth and fifth labials. Measurements in mm.: Total length, 695; tail, 124; width of head, 18; length of head, 22. Variation: No. 32043 is a light fawn-colored specimen bearing a series of black spots with white centers, the outer portion of the spots detached and forming a second row of smaller spots on each side, likewise ocellated. The outer rows and entire ventral surfaces immaculate. There are eleven spots on body and four on the tail. The tail is distinctly triangular in cross-section and terminates in a stiff sp’ne. The ventra's are 169; subcaudals 53; scalerows 21. Variation in ventral scale-count, in material at hand, is 161 to 185; subcaudals, 42 to 56. Most male specimens have the tail stif- fened. All (except one specimen with 19), have 21 rows of scales about the middle of the body. Distribution: This is one of the commoner snakes of Thailand. It has been found in the following provinces: Chaiyaphum, Loei, Sara Buri, Chon Buri, Nakhon Phanom, and Sakon Nakhon. It 768 THe UNIVERSITY SCIENCE BULLETIN probably occurs widely in the plateau areas and the lowlands of cen- tral, eastern and southeastern Thailand. It has been taken in Bang- kok and the northern part of the peninsula. Remarks: 1 have found these snakes very unpleasant to handle as they bite readily. Even when the head is held in the hand the serpent has a propensity of pushing out the back part of the maxil- lary sidewise and biting, the posterior teeth being the ones that seem to cut the flesh to produce much bleeding. Malcolm Smith remarks that “In disposition most of the species (of Oligodon) are quiet and inoffensive. O. cyclurus however, in my experience, is a most vicious creature.” Oligodon cyclurus superfluens subsp. nov. (Fig. 37) Type: No. 77, ¢ Doi Suthep, near Chiang Mai (city) at low elevation (150m.); Edward H. Taylor, Coll., Sept. 6, 1959. Diagnosis: A large series of tiny black marks on edges of dorsal scales (usually not more than two edges of a scale dark) the marks tending to form series of transverse, sometimes angular marks, the angles directed forwards, then laterally they may form a second angle. One or two alternating scalerows may have scales edged with cream or yellow-cream, the color extending on interstitial skin. Head markings similar to cyclurus cyclurus; venter immaculate. Scalerows, 21 at middle; ventrals, 179, subcaudals, 47; head wide; tail rigid ending in a sharp point. Description of type: Rostral somewhat thickened, about as high as wide; portion visable above triangular, its length above, at least as long as its distance from frontal, and tending to displace inter- nasals to a somewhat oblique position; suture between internasals a little shorter than that between prefrontals; latter scales about twice as wide as long; frontal pentagonal; shorter than its distance from tip of snout, shorter than parietals; nasal divided, anterior part the larger; loreal quadrangular, higher than wide; two pre- oculars, lower (subocular) between third and fourth supralabials; two postoculars, temporals, 2 + 2 + 3, upper anterior touching both postoculars, lower anterior separated from them; supralabials eight, fourth and fifth entering orbit; nine infralabials, four touching first chinshields which are one-and-three-fourths times size of second pair. SERPENTS OF THAILAND 769 Fic. 37. Oligodon cyclurus superfluens subsp. nov. Type. No. 4, Doi Suthep (at low elevation, near base), Chiang Mai, Chiang Mai province, northern Thailand. Actual total length, 603 mm.; tail, 90 mm. Scales smooth, the scalerows: 21, 21, 17; ventrals, 179; anal, single; subcaudals, 47. Tail triangular in cross-section. Color in life: Above general body color light brownish gray with a series of about 70 transverse black marks made up of the black edges of scales; many marks are angular mesially, the angle di- rected backwards, the black marks not passing below second scale- rows. Top of head, a little darker than on sides. 30—1367 770 THE UNIvERSITY SCIENCE BULLETIN A long chevron-shaped mark on nape, its rounded apex terminat- ing on frontal, its posterior branches reaching back to level of ninth ventral scale; two oblique bars across snout extending from front edge of frontal to rostral, narrowing, passing through eye to mouth and crossing fifth and sixth supralabials; no black marks on follow- ing labials; many transverse scalerows without black edges, but with cream marks instead, also confined to scale-edges, but extend- ing onto interstitial skin. Towards posterior part of body a sug- gestion of a median whitish streak beginning at level of vent, termi- nating on the sharp-pointed, semi-rigid tail; sides of tail largely unicolored, with a dorsolateral trace of black marking on a slightly darker ground color; a small dark dorsal blotch an inch in front of vent level. . Measurements: Total length, 603, tail, 90; width of head, 16, length of head, 214. Remarks: This species bears a remarkable superficial resemblance to Oligodon cinereus multifasciata that occurs in the same locality. There is however in this form an immaculate venter. The head- markings are wider and more strongly defined such as is typical of those in cyclurus. The tail is proportionally longer, triangular in cross-section, not rounded above and flat below. Oligodon purpurascens purpurascens (Schlegel ) (Fig. 38) Xenodon purpurascens Schlegel, Essai sur la physionomie des serpents, vol. 2, 1837, p. 90, pl. 3, fig. 18, 14 (type-locality, Java? ). Simotes purpurascens: Boulenger, Catalogue of the snakes in the British Museum, vol. 2, pp. 218-219 (part.); A vertebrate fauna of the Malay Peninsula. . . Reptilia and Batrachia, 1912, pp. 148-149. Diagnosis: Rostral not distinctly thickened, length of part visible above less than its distance from frontal; scales in 19 rows (21); below many ventrals have quadrangular dark spots usually alternat- ing with ventral having no spots; subcaudals numerous, 53-56 in peninsular specimens. There are certain populations of purpurascens worthy of sub- specific designation. Several of these have been named, but a revision of the genus will be necessary to bring about a proper understanding of the group as a whole. Description of species (from No. 3894 Na Pradoo, Pattani, Thai- land): Rostral not or scarcely thickened, not or scarcely higher than wide; part visible above triangular, its length less than its distance from frontal; internasals about twice as wide as long, SERPENTS OF THAILAND mad Fic. 38. Oligodon purpurascens purpurascens (Schlegel). No. 3894, Na Pradoo, Pattani province, Southern Thailand. Actual total length, 2922 mm.; tail, 38 mm. somewhat diagonally placed, their mutual suture a little longer than that between prefrontals; frontal much longer than wide (3.5 x5 mm. ), a third longer than its distance to tip of snout, longer than parietals; parietals a little longer than wide; anterior nasal larger than posterior; loreal higher than wide; one preocular and a very small subocular; supraocular about half greatest width of 1 frontal; two postoculars; temporals, 2 +. ——— ; eight supralabials, 1+1 fourth and fifth bordering orbit of eye; mental narrower than rostral; 8-9 infralabials, 3-4 touching first chinshields which are more than double size of second pair; latter separated from first widened ventral by a single pair of scales. Scale formula: 22 (occiput), 19, 19, 15; ventrals, 173; subcaudals, 56; anal, single. M2, Tue UNIVERSITY SCIENCE BULLETIN Color: Above light gray-cream with a series of 19 transverse spots, (four confined to tail), the spots widening in two places dorsally, much narrower laterally and edged with black, separated from each other by an area of ground color with three thin trans- verse dark lines; a row of somewhat quadrangular brown spots on outer part of ventrals and subcaudals, except on neck and posterior part of tail; two ventrals with brown spots alternating with two lacking spots, or each alternate ventral bearing a spot; a curving band of brown on snout passing through eye onto fifth and sixth labials; a large arrow-shaped dark mark on nape bordered with deep brown or black, terminating on middle of frontal, deeply emarginate behind, the posterior parts terminating on each side on outer scalerow at level of ninth ventral; diagonal line from parietal across jaw-angle barely indicated; venter ivory; some small marks on tip of snout below. Measurements in mm.: Total length, 222: tail, 38; width of head, 9: length of head, 12.2. Variation: Boulenger (A vertebrate fauna of the Malay Penin. 1912) gives the variation in ventrals as 160 to 210; subcaudals, 40 to 60. There may be two suboculars and two preoculars; at middle of body scales are in 19 (21) rows. Distribution: The described specimen represents the first one reported in Thai territory. It was collected by Nai Prayoon Kananu- racks. Outside of Thailand it occurs in Malaya, Singapore and in many islands of the Indo-Australian Archipelago. Remarks: Mr. Tweedie (1954) makes the following comments on the coloration in life of the young and adults: “Brown or pur- plish above with black-edged darker cross-bars which are broadest dorsally and may have their margins wavy; between these broad cross-bars other ill-defined narrower ones are usually present. Under surface pink or yellow with square dark spots on the ends of alternate, or more widely separated, ventrals.” . . . “The whole colour pattern becomes gradually more obscure with age, but the young are quite brilliantly coloured, the ground being light brown with scattered red spots, the main cross-bars wholly black, and the underside coral pink.” He states that the scales are in 19 or 21 rows and that the serpent reaches a length of about three feet. It is quite common in wooded country, SERPENTS OF THAILAND 773 Oligodon quadrilineatus (Jan) (Fig. 39) Simotes quadrilineatus Jan, Nouv. Arch. Mus. Paris, vol. 2, p. 7; Jan and Sordelli Iconographie général des ofidiens, tome. 1, livr. 12, 1865, pl. 4, fig. 3 (type-locality, Siam). Simotes taeniatus: Boulenger, Catalogue of the snakes in the British Museum, 2nd ed., vol. 2, 1894, pp. 227-228 (part.); M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 2, 1914, pp. 98-99. Holarchus taeniatus taeniatus: Cochran, Proc. U. S. Nat. Mus. vol. 77, 1930, p. 28 (Bangkok, Nonthaburi, Korat, Pak Jong, Nong Mong). Oligodon quadrilineatus: M. Smith, The fauna of British India, Ceylon, and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, sec. 1943, p. 210; Taylor and Elbel, Univ. Kansas Sci. Bull., vol. 38, pt. 2, 1958, pp. 1140-1142, fig. 27. Diagnosis: Two dorsal and two lateral brown stripes; dorsal lines separated by a lighter, partly yellow line; scale-formula: 19, 19, 15; no black stripe crossing rostral; four anterior labials on each side; no black spot at base of tail or near tip; a blackish chevron on nape and two oblique temporal marks; a wide black mark on snout, narrowing and crossing eyes to lip; reddish on venter. Description of species (from No. 667 Bangkok): Rostral a little higher than wide, triangular part visible above as long as inter- nasals; latter angular anteriorly, wider than long, their common suture about two thirds as long as that between prefrontals; latter scales one-and-a-half times as wide as long, their suture with frontal forming a straight line; frontal longer than its distance from tip of snout, only slightly shorter than parietals; nasal divided, anterior part much the larger; loreal longer than high; one preocular, two 1 postoculars; temporals, 1 + 1 + —, eight supralabials, fourth bor- 2 ders orbit, sixth and seventh largest; nine infralabials, four bordering first chinshelds which are twice size of second. Scale-formula: 22 (occiput), 19, 19, 15; scales smooth without apical pits. Ventrals, 157; subcaudals, 41; anal, single. Color in life: Body and head gray with two longitudinal median dorsal blackish stripes to tip of tail, separated by a narrow yellowish- white line passing through median scalerow, but not covering whole scales. Edges of the brown scales yellow, as is also interstitial skin; a lateral line on fourth and fifth scalerows terminating at level of vent; head with a transverse mark crossing frontal and back part of prefrontals, passing through eye and across fifth and sixth suprala- bials; a spot on frontal; two diagonal temporal stripes arising at edge THE UNIVERSITY SCIENCE BULLETIN ig. i ie: ‘ sow" Fic. 39. Oligodon quadrilineatus Jan. KUMNH No. 31437, Banlat Ban Kaeng, Phukheio, Chaiyaphum province, Thailand. Actual total length, 354 mm. SERPENTS OF THAILAND TAS of supraoculars and terminating on sides of neck; an arrow-shaped mark on neck, terminating on back of parietals connected behind with two dorsal stripes. Venter laterally angulate but bluntly so; lateral part of ventrals cream or ivory; chin and throat dirty ivory, gradually growing pink, then reddish to near tip of tail, with numerous ventrals bearing more or less quadrangular spots, form- ing two rows, the spots sometimes alternating, sometimes on the same ventral (rarely connected ) and usually separated by a ventral lacking spots; ventral part of the tip of tail white. Most gray scales of body, under a lens, show some black along scale edges and on the interstitial skin. Table of data on Oligodon quadrilineatus (Jan) NUMBER 667 672 34166 34168 34169 Total length....... 375 375 388 402 342 Mev ilicte eine. ce. 63 64.5 67 66 62 WATvall pensar 1 16 1 1 1 Ventrals........... 157 15 159 157 157 Subeaudals........ 41 44 43 43 44 Supralabials....... 8-8 8-8 8-8 8-8 8-8 Infralabials........ 9-9 9-9 9-9 9-9 9-9 Preocular.......... 1-1 1-1 1-1 1-1 1-1 Postoculars........ TD) 2-2 2-2 2-2, 2-2 Temporals......... 1+1+4 1+1+4 1+1+4+4 1+1+4 1+1+4 Seale-formula...... KOA Non OPO ono Oa OT LO lon lO oO ale Distribution: This species is found commonly at Bangkok in cen- tral Thailand. Specimens are known from Chaiyaphum, Nakhon Ratchasima and Sara Buri. It is presumably endemic in Thailand. Remarks: The species apparently, in the region of Bangkok main- tains its identity although the closely related taeniatus occurs also. They are both small forms, having a reddish venter with quad- rangular spots. The latter species differs in the number of scale- rows at the middle of body (19 instead of 17), there are no large blotches on the tail, no black band bordering the front of the upper lip, and the bar on the snout is much narrower. There is some confusion in the literature. Gunther's original description of taeniatus stated there were 19 scale rows, but later this was corrected to 17. Boulenger, 1914 (p. 70) has overlooked the correction and proposes a new name needlessly. The specimens listed in the table are from Bangkok. 776 THe UNIVERSITY SCIENCE BULLETIN Oligodon barroni (Smith) (Figs. 40, 41) Simotes barroni M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, No. 1, 1916, pp. 46-47, pl. —, fig. 4, 4a (type-locality, Hup Bon, E. of Sriracha, Chon Buri province, Thailand). Holarchus taeniatus candaensis Bourret, Bull. Gen. Instr. Pub. Hanoi, May 1934, p. 173 (type-locality, Canda, near Nhatrang, §. Annam (fide M. Smith). Oligodon barroni: M. Smith, The fauna of British India, Ceylon, and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, Dec., 1943, pp. 210-211, fig. 66 (dorsal pattern ). Diagnosis: A small species (380 mm.) with 10-12 dark-brown transverse ventral bars tending to be confluent with small lateral brown spots; these separated by three more or less distinct dark narrow ‘transverse markings. Near outer part of ventrals a light Fic. 40. Oligodon barroni M. Smith. No. 1719 young, Muak Lek, Friend- ship Highway, Sara Buri, province, Central Thailand. Actual total length, 222 mm.; tail, 36.3 mm. SERPENTS OF THAILAND TUT line, with two ventral rows of quadrangular black spots. A broad, somewhat heart-shaped nuchal spot with a forward arm; a diagonal stripe across angle of mouth closely approximating nuchal spot, all more or less outlined with gray white; a faint pair of dorsal brown- ish lines with irregular spotting, and a faint lateral line, at least posteriorly; one anterior temporal; seven (eight) supralabials, two bordering orbit; 135-160 ventrals; venter coral-red, fading to white in preservative. Description of species (from No. 33245 Ang Hin, Chon Buri): Rostral as high as wide, not thickened, well visible above, angular mesially; internasals wider than long, their common suture less Fic. 41. Oligodon barroni M. Smith. No. 341494, Ang Hin, Chon Buri province, southeastern Thailand. Actual total length, 372 mm.; tail, 54 mm. 778 THE UNIVERSITY SCIENCE BULLETIN than one half their width; prefrontals about twice as wide as long; frontal pentagonal, front edge a straight line, sides not parallel, its length equal or slightly greater than its distance from tip of snout, shorter than parietals; nasal divided, anterior part a little larger than posterior; loreal quadrangular, longer than high; preocular narrowed above not reaching top of head; two subequal postoculars; 1 temporals, 1 + —; supralabials, 7-8, third and fourth (fourth and fifth) entering orbit; infralabials, 9-9, four in contact with first chinshields which are nearly double size of second pair. Scales smooth, the formula: 17, 17, 15; ventrals, 152, anal, single, subcaudals, 46: ventrals angular laterally their outer parts turned up on sides. Color in life: Above generally light brown with a faint trace of two dorsal stripes of darker brown and a still fainter lateral stripe; head with a dark brown transverse crescent across frontal-prefrontal area passing across eyes to terminate on fifth (sixth) labials; a small dark spot on the following labial; a diagonal stripe from parietal to third ventral; a heart-shaped nuchal spot, its apex on posterior part of parietals; all head-spots edged with gray-white; on body eleven rather distinct dark brown bars, often confluent with a lateral series of smaller spots, and mesially indented, in some cases dividing the spot; between these median spots three more or less definite narrow transverse marks; tail with four transverse bars, but intervening marks absent; chin whitish; venter pink to coral red with two rows of brown quadrangular marks, extending on to subcaudal region almost to tip of tail; rostral gray-white with two small dark vertical marks. Measurements in mm.: Total length, 364; tail, 70.5; width of head, 8; length of head, 13. Distribution: Only a few specimens of this species have been taken. Malcolm Smith who described the form in 1916 had three specimens from Hup Bon, Sriracha, Chon Buri. It has been taken on the island Koh Lam near Sriracha, and at Muak Lek altitude 199 ft. in Sari Buri province. My two specimens are from Ang Hin, Chon Buri. A third specimen in the collection is possibly also from this place. Variation: The number of labials is variable. Smith reports three having seven on one side, eight on the other, while a single speci- men has eight on each side. He reports one specimen lacking the SERPENTS OF THAILAND 779 larger dark spots on the anterior part of the body. In the specimen I figure, the three anterior dark spots are scarcely larger than the intervening marks. A young specimen from Muak Lek, has a count of 160 ventrals. The dorsal brownish stripes are not mentioned in the type de- scription. However the excellent figures show these as well as an indistinct lateral line. Oligodon taeniatus (Gunther ) Simotes taeniatus Giinther, Proc. Zool. Soc. London, 1861, p. 189; The reptiles of British India,* 1864, p. 216, pl: 20, fig. A ‘(type- locality, Cambodia) Boulenger, Catalogue of the snakes in the British Museum, 2nd. ed.; vol. 2, 1894, pp. 227-228 (part.); Boettger, Ber. Senckenb. Nat. Ges., 1894, p. 134, (part. states that taeniatus = quadrilineatus); Barbour, Proc. New England Zool. Club, vol. 4, 1909, p. 70; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 1, p. 98 (Bangkok). Simotes taeniatus mouhoti Boulenger, Journ. Nat. Hist. Soc. Siam, vol. 1, 1914, p. 70 (type-locality, “Gamboja” = Cambodia). Holarchus taeniatus mouhoti: Cochran, Proc. U. S. Mus., vol. 77, 1930, p. 29; Bourret, Les Serpents de Indochine, 1936, p. 247. Oligodon taeniatus M. Smith, The Fauna of British India, Ceylon, and Burma . . Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, p. 208-210, figs. 64, A and B; fig. 65. Diagnosis: A small species with a very broad band across snout, narrowing across eyes and labials; an elongate frontal spot extend- ing onto parietals; two oblique temporal stripes reaching down onto ventrals; a symmetrical blotch on nape reaching back of parietals; reddish below except on chin and neck; numerous quadrangular black spots on venter some fusing and making bars, not arranged in rows; a median dorsal yellowish line; two dorsal darker stripes ex- tending to tail and a lateral stripe stopping at level of vent; a spot at base and one near tip of tail; a continuous black band across anterior labials and rostral; two preoculars; eight supralabials; 17 scalerows. Description of species (from No. 609, Hua Hin, Prachuap Khiri Khan): Rostral as high as broad, the part visible above equal to about half its distance from frontal; internasals somewhat angular anteriorly a little wider than long; prefrontals about twice as wide as long; frontal subpentagonal longer than its distance from tip of snout, a little shorter than combined parietals; nasal divided, anterior part the larger; loreal quadrangular longer than high; two preoculars, lower small, partly between third and fourth * The original description states 19 scale rows. In the second report Giinther corrects this to 17 scalerows. 780 THe UNIVERSITY SCIENCE BULLETIN supralabials; two postoculars; first temporal small, touching both 1 postoculars; temporal formula, 1+ 1+ —,; eight supralabials, 2 fourth and fifth bordering orbit, last four large; eight-nine supra- labials, four bordering first chinshields which are a little less than twice size of second pair; scales smooth in 17, 17, 15 rows, without pits; ventrals, 148; subcaudals, 43; anal, single. Color in life: Grayish above with two light-brown median stripes covering little more than one scalerow, separated by a narrow lighter line, the scales of which have yellow spots or edges, and much of the skin between scales also yellow. Head, anteriorly with a broad band across snout, narrowing across eyes, and extending posteriorly across three supralabials; anterior supralabials white with a band beginning on rostral extending along lip to fifth labial but not onto it; an oblique temporal band extending from parietals under neck onto ventrals, but not forming a connection; a symmetrical black blotch on neck terminating on back of parietals; an elongate spot on frontal; all head-marks outlined in whitish; venter generally pink with quadrangular spots, few anteriorly and practically none on lat- ter part of tail, some spots joining to form bars on venter; chin ivory-white. Measurements in mm. (of Nos. 608 ¢, 609 2 ): snout to vent, 212, 184; tail, 38, 22.5; width of head, 6, 5.8; length of head, 10, 9.8. Variation: The second specimen measured is a young female. The ventrals are 157, the subcaudals, 32. The tail is complete. In color and general squamation the two specimens agree almost per- fectly. They are from the same locality. Malcolm Smith, loc. cit. records the total length of one male specimen as 340 mm.; the tail 60mm. A second female specimen as 330 mm. the tail 45. The variation in ventrals is 146 to 169; in subcaudals, 30 to 47. Distribution: The species is not uncommon at Bangkok where it occurs with Oligodon quadrilineatus. Specimens have also been taken in “Siam between lat. 12° and 16° N”; Prachuap Khiri Khan. Outside of Thailand the species is known in Cambodia and Cochin China. Remarks: My specimens were found under piles of recently cut grass on the golf course at Hua Hin. SERPENTS OF THAILAND 781 Addenda: Malcolm Smith, loc. cit., 1943 gives the range of this species in Thailand as “Siam between lat. 12° and 16° N.” Two recently acquired specimens show the range extending somewhat farther south. EHT-HMS No. 3490 is from the city of Songkhla, collected by Dr. Robert Elbel; and EHT-HMS 3908, Na Pradoo, Pattani, collected by Nai Prayoon Kananuracks. In these the median blackish figure on the neck reaches to the frontal but does not connect with the two blackish diagonal bars crossing jaw-angles and bending across the throat forming a con- tinuous unbroken bar on throat in one, while in the other the mark is interrupted mesially. The Songkhla specimen has only about ten scattered quadrangular spots; the Na Pradoo specimen is al- most immaculate for nearly eight centimeters and two centimeters from tip of tail, while the remainder is heavily spotted with black bars or quadrangular black spots alternating. The venter of the Songkhla specimen was salmon pink. That of the Na Pradoo speci- men, now cream, may have been so colored in life. The ventral-subcaudal counts are respectively, 146, 42, and 155, 42. Na Pradoo is 7° North latitude. Oligodon joynsoni (Smith) Simotes longicauda joynsoni M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, no. 4, Dec. 1917, pp. 276-277 (type-locality valley of the Maa Yome, Muang Ngow, Northern Thailand). Holarchus longicauda: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 4, no. 4, July 25, 1922, p. 208 (part.). Oligodon joynsoni: M. Smith, The fauna of British India, Ceylon, and Burma ._. .. Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, p. 218, fig. 72 (dorsal pattern ). Diagnosis: A medium-sized species (760 mm. in length); bodily configuration similar to cyclurus; eight supralabials; a subocular and a preocular present; scales smooth in 17 rows at middle of body, 15 shortly behind middle; ventrals feebly angulate; dark purplish brown with strong black reticulations forming more or less distinct but broken crossbars, each alternate one with a transversely placed black spot; venter red in life, heavily marked with rectangular black ‘spots; a narrow chevron terminating on head; a crescent spot across eyes and snout; two temporal bars. Ventrals, 187-197: subcaudals, 40-50. 782 THE UNIVERSITY SCIENCE BULLETIN Description of species (data from type description ): Rostral bent over tip of snout, portion visible above as long as its distance from frontal; suture between internasals shorter than that between pre- frontals; frontal longer than its distance from end of snout, as long as parietals; loreal slightly longer than deep; one preocular (usually one presubocular) and two postoculars; temporals, 1+ 2 (or 1+1+ 2, anterior small); eight supralabials, fourth and fifth bordering orbit; five infralabials touching first chinshields which are twice as large as posterior pair; nasals divided. Scale-formula: 17, 17, 15, 15, the reduction taking place just posterior to mid-body; ventrals, 190; anal, single; subcaudals, 47; maxillary teeth, 12, steadily increasing in size posteriorly; palatine teeth, 7, pterygoid, 14. Color: Purplish brown above with ill-defined broken bands across back, every alternate one enlarged into a dorsal blotch, about 50 altogether on body and tail; below yellowish white with rather large black quadrangular spots not reaching outer lateral part of ventrals; a large dark-brown crescentic band in front, passing across prefrontals and through eyes; an oblique temporal streak present, and a narrow chevron on nape terminating on frontal. Measurements in mm.: Total length, 760; tail, 105. Variation: A second specimen differs in having the frontal shorter than the parietals, a subocular present, and only seven supralabials on left side (due to fusion of the second and third). The ventrals are 187, the subcaudals 50. It is light brown above, the dorsal blotches being scarcely enlarged. Below it is uniform yellowish white. Other specimens show that the subocular below the pre- ocular may be absent and there may be one or two anterior tem- porals. Ventrals vary between 187 and 195, the subcaudals be- tween 43 and 50. The venter is always reddish in life. Remarks: Smith himself seems to have had some doubts as to the validity of this species. In 1922 he referred the form to Oligodon longicauda. He states: “the difference in colouration be- tween them was far less than [ gathered from the description, so much so that mine can no longer be recognized as a form apart.” However in 1943 he recognizes joynsoni, not as a subspecies of longicauda, but as a distinct species. Not having examined ma- terial of this species I accept Smith’s 1943 decision as to its status. Distribution: Known only in the type-locality and from Me Wang, Northern Thailand. SERPENTS OF THAILAND 783 Genus Liopeltis Fitzinger Liopeltis Fitzinger, Systema reptilia, 1843, p. 26 (type of genus, Herpetodryas tricolor). Gongylosoma Fitzinger, ibid., p. 25 (type of genus, Coronella baliodeira. Ablabes Duméril, Mem. Acad. Sci. Paris, vol. 23, p. 454 (type of genus, Coronella balioderia). Diagnosis: Small snakes, scales in 13-17 rows around body (18 in Thai forms), not reducing posteriorly; apical pits lacking; one preocular, one or two postoculars. Maxillary teeth, 24-28, about equal in size in Thai species. Head not or scarcely distinct from neck in Thai species. Two species are recognized in Thailand: scriptus and baliodeirus. The latter is represented by a recently described subspecies b. cochranae. The two forms may be separated by the following key: 1. Ventrals, 115-137; subcaudals, 58-75; eight supralabials; several dark bands around anterior part of body; two preoculars; two postoculars, baliodeirus cochranae 2. Ventrals, 117-145; subeaudals, 89-103; supralabials eight; a narrow inter- rupted band across occiput, bordered with black, followed by a broad somewhatedarker saddle=shaped blotch.:.....,....-. -4-. 3). scriptus Liopeltis baliodeirus Boie in Boie Coronella baliodeira Boie in Boie, Isis, 1827, 539 (type-locality, Java): Schlegel, Essai sur la physionomie des serpents, vol. 2, 1837, p. 64, pl. 2, figs. 9, 10 (including a “var.” in Sumatra); Cantor, Journ. Asiat. Soc. Bengal, vol. 16, pt. 2, 1847, p. 913. Ablabes baliodeirus: Duméril, Bibron and Duméril, Erpétologie générale . . . 1854, p. 313; Giinther, Catalogue of the colubrine snakes in the collection of the British Museum, 1858, p. 29; The reptiles of British India, 1864, p. 224; Flower, Proc. Zool. Soc. London, 1899, p. 673; Boulenger, A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 152; Catalogue of the snakes . . . vol. 2, 1894, pp. 283-284. Diadophis baliodeirus: Jan, Arch. Zool. Anat. Fis., vol. 2, 1863, 263; Jan and Sordelli, Iconographie générale, livr. 15, pl. 5, fig. 4. Gongylosoma baliodeirum: Sworder, Singapore Nat., no. 2, 1922, p. 65; ibid., no. 3, 1921, p. 22 (Singapore). M. Smith, Bull. Raffles Mus., no. 3, 1930 pp. 56-57. It is obvious from the variation reported, that several subspecies of this snake exist. Schlegel recognized a “var.” in Sumatra. (Schle- gel, Atlas, on—“Tableau servant a illustrer la répartition des ser- pent a la surface du globe”). J. A. Fisher, described Ablabes baliodirus var. cinctus (Abh. natur. Ver Hamburg, vol. 9, 1886 (1885), p. 8, pl. 1, fig. 2, from Nias). I have not seen specimens of this species from Malaya. It is known in Penang, Province Wellesley and on Bujong Malacca, Perak (fide Boulenger, 1912). 784 THE UNIVERSITY SCIENCE BULLETIN Whether any of these specimens are the same as the form de- scribed here, remains to be ascertained. Liopeltis balioderius cochranae Taylor (Fig. 42) Liopeltis baliodeirus cochranae Taylor, Univ. Kans. Sci. Bull., vol. 43, 1962, pp. 261-263, fig. 18 (type-locality, Khao Soi Dao, Thailand). Diagnosis: Thirteen scalerows around body; nostril between two nasals fused above nostril; head scarcely distinct from neck; two preoculars, two postoculars; eight supralabials, the fourth and fifth border eye; 24-25 maxillary teeth; neck with eight or nine black bands (reaching to ventrals) all except first, third, and fourth in- terrupted on median dorsal line. Description of type: Head a little wider than neck; rostral not twice as wide as high, narrowly visible above on snout; internasals wider than long, much smaller and narrower than prefrontals; lat- ter much broader than long, laterally touching loreal, posterior nasal, and upper preocular; frontal shield-shaped, longer than its distance from tip of snout, shorter than parietals; nostril between two nasals that are fused above nostril, the posterior nasal largest, distinctly higher than anterior nasal; a small loreal; two preoculars, upper not reaching upper surface of head; supraoculars large, not as wide as frontal; two postoculars; a single large anterior temporal followed by two superimposed temporals much shorter than anterior; eight supralabials in the following order of size: 3, 2, 1, 6, 4, 5, 8, 7; the fourth and fifth enter orbit; mental triangular, first labials in contact behind it; seven infralabials, first three touch first chin- shields which are much shorter than second pair. Scalerow-formula: 17 (occiput), 13, 13, 13: scales smooth, with- out apical pits. Ventrals, 118; subcaudals, 72; anal, divided. Color: Head nearly uniform light brown; on side of head very slight dark marks evident along sutures of some labials and a fine dark rim on edges of scales marking outline of orbit; on occiput the brown tends to become blackish in front of first cream bar which is angulate anteriorly and almost broken on side; this followed by a series of eight or nine blackish bars (some of which are broken by a middorsal light line), and separated by dull cream or fawn- colored bars; dark and light bars becoming less distinct posteriorly where they fade into the uniform, finely reticulated pattern of violet brown, each scale with a diamond-shaped light, gray-lavender cen- ter,—a pattern that continues to tip of tail. Outer edge of ventrals SERPENTS OF THAILAND 785 Fic. 42. Liopeltis baliodeirus cochranae Taylor. Type, USNM No. 94826 (6343), Khao Soi Dao, Thailand. Actual total length, 374; tail, 118. (From Taylor, 1962. ) 786 THE UNIVERSITY SCIENCE BULLETIN and subcaudals with a violet-brown line separating the uniform whitish color of underside of head, body, and tail from the dorsal coloration. Measurements in mm.: Snout to vent, 256; tail 118; total length, 374; width of head, 9.1; length of head, 14. Remarks: Another species of the genus, Liopeltis scriptus (Theo- bald), also occurs in peninsular Thailand and specimens have been taken in the Nakhon Si Thammarat Mountains (Khao Luang and Ronpibon ), and on Pulau Panjang (Island) of Phuket. The differences are considerable in these two forms, both having 13 scalerows about the body. In b. cochranae ventrals are fewer, 118 compared with 122-145; the subcaudals fewer, 72, compared with 87-103. There are two instead of three supralabial scales bordering orbit; the nasals are partly fused (two separate), and the anterior temporal is wider and somewhat shorter than in L. scriptus. The subspecies in known only from the type-locality. Liopeltis scriptus (Theobald) (Fig. 43 ) Ablabes scriptus Theobald, Journ. Linnean Soc., vol. 10, no. 41, Zoology, 1868, p. 42 (type-locality, Martaban, Burma); Catalogue of the reptiles in the museum of the Asiatic Society of Bengal (Extra no. of Jour. Asiat. Soc. Bengal, 1868, p. 49); Boulenger, The fauna of British India . Reptilia and Batrachia, TO, pp. 305-306; Catalogue of the snakes in the British Museum, 2nd ed. vol. 2, 1894, p. 284. Liopeltis scriptus: Wall, Journ. Bombay, Nat. Hist. Soc., vol. 29, 1924, p. 864. Gongylosoma scriptum: Cochran, Proc. U. S. Nat. Mus., vol. 77, pt. 2, 1930, p. 30; M. Smith, Bull. Raffles Mus., no. 3, 1930, pp. 56-57 (“Petchaburi District” Thailand; Telok Poh, Phuket; Khao Luang, Nakhon Si Tham- marat); The fauna of British India. . . Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, pp. 186-187, fig. 57. Diagnosis: Small snake, reaching about half meter in length; scales in 13 smooth rows throughout body; eight supralabials, third and fourth entering eye; anterior temporal longer than secondary temporals; a narrow black-edged broken cream band across occiput, preceding a dark neck band; supralabials cream with some black along certain sutures. Description of species (from No. 35587 Ronpibon, Nakhon Si Thammarat): Rostral not twice as wide as high, visible above; internasals moderately large touching nasals laterally, smaller than prefrontals; latter nearly twice as wide as long, touching posterior nasal, loreal, and preocular laterally; frontal longer than its distance from tip of snout, shorter than parietals; nasal divided, anterior part angular and a little the larger, posterior part rounded behind; SERPENTS OF THAILAND Fic. 43. Liopeltis scriptus (Theobald). No. 35587, Ronpibon, Nakhor Si Thammarat, Thailand. Actual total length, 355 mm.; tail, 152 mm. 788 THe UNIVERSITY SCIENCE BULLETIN loreal small, a little higher than long; single preocular high and narrow, but not reaching top of head; supraoculars little more than half width of frontal; two postoculars both touching anterior tem- poral; temporals, 1 + 2, anterior longer than posterior; eight supra- labials, third, fourth, and fifth border orbit; seventh distinctly the largest; diameter of eye equal to its distance from rostral; eight infralabials, last very small, first four border first chinshields which are shorter than second pair. Scales in 13 rows around body; ventrals, 122; anals, 2; subcaudals, 103 (104). Color: Above generally grayish, scale edges darker brown; two outer scalerows with edges of scales less distinctly marked except on scales bordering third scalerow, where edges are a little more heavily marked, leaving a slightly darker line the length of body. Head gray to brownish gray above, anterior part dimly flecked with darker; side of snout light fawn with dark marks on rostral, nasal sutures, and on first and second supralabials; a narrow black ring around eye, widening considerably on labials below eye; a cream line from fifth and sixth labials pushes up back of eye partially bordered behind by black; a black-edged cream line from eighth supralabial reaches to or nearly to parietal; a dark band four or five scales wide on neck, bordered narrowly by cream posteriorly; venter and subcaudal region cream. Measurements in mm.: Snout to vent, 203; tail, 152; total length, 355; width of head, 6.3; length of head, 12. Variation: USNM No. 76103 is also from Nakhon Si Thammarat. It has 133 ventrals and 98 subcaudals. The head and body measure 240 mm., the tail 160 mm.; it differs in having two preoculars and the light transverse mark behind the dim nuchal band is scarcely discernible; a few tiny punctate black flecks present for about two inches behind head. The ventrals are known to vary between 122 and 145, the sub- caudals, 87 to 103. Distribution: In Thailand the species is known from Sai Yoke, Kanchanaburi province; Khao Luang and Ronpibon, Nakhon Si Thammarat; Pulao Panjang (Island), Phuket. It has not been reported outside Thailand except in Burma (Martaban ), the type-locality. Genus Dryocalamus Gunther Dryocalamus Giinther, Catalogue of the colubrine snakes in the British Museum, p. 112 (type, D. tristrigatus ). SERPENTS OF THAILAND 789 Diagnosis: Eye large with a vertical pupil; scales in 13 or 15 sub- equal rows, normally imbricating, with apical pits present. Hypa- pophyses lacking on posterior body vertebrae. Maxillary bone curving inward anterior to palatine; teeth on maxillary 8 to 10, in- creasing in size posteriorly. The genus ranges from India throughout southeastern Asia, Malaya to the Philippines. One species is known from Thailand. Dryocalamus davisonii (Blanford ) (Fig. 44) Ulupe davisonii Blanford, Journ. Asiat. Soc. Bengal, vol. 47, 1878, p. 128 (type-locality, Nawlabu Hill, East of Tavoy, Burma); Proc. Zool. Soc. Lon- don, 1881, p. 221. Hydrophobus davisonii: Boulenger, Fauna of British India, including Ceylon and Burma; Reptilia and Batrachia, 1890, p. 299. Dryocalamus davisonii: Boulenger, Catalogue of the snakes in the British Mu- seum, vol 1, 1893, p. 372; Wall, Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 616; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, 1914, p. 93; Bourret, Les Serpents de lIndochine, 1936, p. 168; M. Smith, Fauna of British India, Ceylon and Burma, including the whole of the Indo-Chinese Subregion; Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 274-275. Taylor and Elbel, Univ. Kan. Sci. Bull., vol. 38, pt. 2, 1958, p. 1044 (listed as D. davidsonii [sic] ). Diagnosis: Nostril in a single nasal; no preocular, loreal entering orbit of eye; scales in 13 rows; black above, banded with white or pale green; bands posteriorly shorter than anteriorly; subcaudals paired. Description of species (from No. 1283, collection of Dr. Boonsong Lekagul): Rostral much wider than high (3.3 < 1.7 mm.), broadly visible above, forming a slight median angle on its upper border; internasals distinctly longer than prefrontals but less wide, nar- rowed anteriorly, anterior and posterior sides parallel; prefrontals with a slight angle on their posterior border, laterally broadly in contact with loreal, narrowly so with nasal; frontal approximately as long as its distance from end of snout, and equal to its distance from posterior level of parietals; parietals in contact for four-fifths of their length; supraoculars widened posteriorly. Nostril in an undivided nasal scale somewhat rectangular in shape; preocular absent; loreal broadly entering orbit of eye; eye large, pupil vertically elliptic; two postoculars; anterior temporals, one on right, two on left side; temporal formula, 1 + 2+2+43 1 (right side), 2-4+2-+ 1-43 (left side); seven upper labials, third 2 and fourth enter orbit; eight lower labials, anterior four touch first THe UNIVERSITY SCIENCE BULLETIN Fic. 44. Dryocalamus davisonii (Blanford). No. 90024, Chalermlaub, Sriracha, Chon Buri province, southeastern Thailand. Actual total length, 444 mm; tail, 74 mm. chinshields which are more than double size of posterior pairs; four pairs of chinshields. Scalerow formula: 17 (occiput), 18, 13, 18; scales small with single apical pits; ventral scales, 233; subcaudal scales, 92; anal, single; ventrals keeled, with a slight notch on each side. SERPENTS OF THAILAND 791 Color: Above with 27 black saddle-shaped blotches narrowed laterally, reaching to ventrals (anteriorly) and posteriorly to outer scalerow or to second or third outer row near vent; the length of saddles anteriorly are equal to 13 scale-lengths; farther back to eight scale-lengths growing gradually shorter to five and four pre- ceding level of vent; a second series, at first small, begins on second row at 14th blotch and continues to vent, growing larger and alter- nating with dorsal blotches. The intervening lighter areas, whitish anteriorly, become cloudy with darker pigment except for a single row of scales bordering dark blotches. The light intervening areas are widened laterally and narrowed dorsally, the edges of both light and dark areas being denticulate; head with a light brown mark covering frontal region to snout with a median posterior mark reaching to first dorsal blotch and to posterior projection passing through eye and onto two temporals; posterior occipital region with a dusting of pigment; chin and venter nearly pure white with a few dark spots on latter fifth; considerable pigment on sub- caudals, dense mesially. Measurements in mm.: See table. Measurements in mm. and scale data on Dryocalamus davisonii NUMBER 1283 1756 Q 900 07 Snout=svemt lenptlhisis ecw Gee ete 536 557 370 Mani lerot hie Miner Oe era Secs ened sa 152 164 74 Mo tailblenothy se wee ih es Wong a oon cas. 688 721 444 WA GIRO RINE AC es apn a ee ee 9 10.3 7 Weme-theotsleadiy eee see ee 14.6 16 12 We rirtereall Seperate ty ye ess oe tes we anne ae gue ies 233 238 238 Sullocau dase sees aoe steerer 92 95 61+ reo culanszer rise cere en once aatw near ae 0) 0 0 Rostocularssee cree ee tr eee 2 1 1 Supnalalsiallsiewe ecco hee et ne cn 7 7 7 Cliimnshieldapairs-cr asta ens acy toe ee 4 4 4 Variation: Known variation in ventral count is between 233 and 255, the subcaudals between 90 and 108. The loreal may enter the eye. Seven or eight supralabials are present. The species reaches a length of 920 mm.; the tail 205 mm. Distribution: In Thailand specimens have been reported between 18° and 11° north latitude. I have taken the species in Chalermlaub, Sriracha, Chon Buri from under the bark of a stump. M. Smith reports two specimens taken from Sapatoom, Bangkok. The species ranges from Burma to Indo-China. 792 THE UNIVERSITY SCIENCE BULLETIN Remarks: This is a lowland species and cannot be considered common. The eggs of the species are remarkably elongate. Smith (1943) reports the size of four laid by a captive specimen as meas- uring 39 mm. * 9 mm. Genus Plagiopholis Boulenger Plagiopholis Boulenger, Catalogue of the snakes in the British Museum, vol. 1, 1893, p. 301, (type of genus, blakewayi). Trirhinopholis Boulenger, Catalogue of the snakes in the British Museum, vol. 1, 1893, p. 419, (type, nuchalis). Diagnosis: Maxillary teeth 16-20, small subequal; nostril between two nasals; eye with a vertical subelliptical pupil; scales smooth or with keels posteriorly on body and base of tail; scales in 15 rows, more or less obliquely arranged, without apical pits. Body cylin- drical, head not distinct from neck; subcaudals all or mostly di- vided; hypapophyses throughout vertebral column; mental touch- ing first chinshields. Four species are recognized. These are distributed through Burma, northern Thailand, southern China, and Indo-China. A single species, Plagiopholis nuchalis, has been taken in Thai- land, this from the area north of latitude 13°. Plagiopholis nuchalis (Boulenger ) (Fig. 45) Trirhinopholis nuchalis Boulenger, Catalogue of the snakes in the British Mu- seum, 2nd Ed., vol. 1, 1893, p. 419, pl. 38, fig. 1, (type-locaiity, Toungyi, S. Shan States); ibid., vol. 3, 1896, p. 612; Ann. Mus, Civ. Genova, ser. 2, vol. 13, 1893, p. 323; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 2, Mar., 1915, pp. 155, 213, (specimens from near border of Thailand, Sai Yoke district); Journ. Bombay Nat. Hist. So., vol. 23, 1915; Wall, ibid., vol. 29, 1923, pp. 466, (Sinbur Kaba, Upper Burma), 612; ibid., vol. 30, 1925, p. 811, (13 specimens, Kachin Hills and Kut Kai, North Shan Hills, Burma); ibid., vol. 31, 1926, p. 561; Taylor, Proc. Acad. Nat. Sci. Philadel- phia, vol. 86, June, 1934, pp. 302-304. Plagiopholis nuchalis: M. Smith, Rec. Ind. Mus., vol. 42, 1940, p. 484; The fauna of British India, Ceylon, and Burma including the whole Indo-Chinese subregion, Reptilia and Amphibia, vol. 3, Serpentes, Dec., 1943, pp. 326- CW: Oligodon evansi Wall, Journ. Bombay Nat. Hist. Soc., vol. 22, 1913, p. 514, fig.,. (type-locality, Toungyi, S. Shan States) and ibid., vol. 27, 1920, p. 175, [= P. nuchalis]. Diagnosis: A small snake (470 mm.) resembling somewhat a species of Oligodon; ventrals, 122-142, subcaudals, 23-30; six supra- labials; loreal present; 15 scalerows, smooth except in anal region; first infralabials separated by first chinshields, which touch mental. Description of species (from No. 33635, Doi Suthep, Chiang Mai, north Thailand): Rostral broader than high, part visible above SERPENTS OF THAILAND 793 Fic. 45. Plagiopholis nuchalis (Boulenger). No. 33635, Doi Suthep, at Forest Station, Chiang Mai, Chiang Mai province. Actual length, 238 mm.; tail, 26 mm. about equal in width to suture between small internasals, which are less than half area of prefrontals; frontal much longer than its distance to end of snout, slightly less than length of parietals; scales bordering temporals and parietals somewhat larger than body scales; each prefrontal is fused to frontal by a narrow isthmus (presumably abnormally but symmetrically); nostril between two nasaks; a quadrangular loreal present, touching one labial; a verti- 794 Tue UNIVERSITY SCIENCE BULLETIN cally elongate preocular; a broad supraocular; two postoculars; temporals, 1 + 2; six supralabials, third and fourth enter eye, fifth much the largest, sixth second in size; six infralabials, three touch- ing large anterior chinshields, which are considerably larger than second pair; first infralabials separated behind mental; scales smooth except near tail in 15 rows throughout, those of two outer rows largest; scales arranged somewhat obliquely (but not conspicuously so ) and keeled in posterior part of body. Diameter of eye much larger (nearly double) its distance from mouth; ventral scales, 126; anal, single; subcaudals, incomplete. Color in life: Somewhat reddish brown with many of the scales edged with black and forming about 56 single or double transverse dark markings across back, which may connect to form a reticulum; ventral scales yellowish, with irregular rows of rather large rectan- gular black marks on each side of venter and in subcaudal region; head yellowish olive with a strong chevron-shaped black spot point- ing forwards and terminating at parietals; chin and throat yellowish as far as fourth ventral; two vertical dark marks below eye. Table of measurements and data on Plagiopholis nuchalis NUMBER 33635 19890 675 9 TO Cale era gta eyes saeed ae ieee ert res eee ce 238% 5 eee 376 Pal ee aaeate see Oe Seen oe ae 26a ee 42 Wirditheo tah ead eegaia sates ie Ares mene 8:2" Slee 9.5 Bengthy ofshead@ ite. 30-0 meee ee WA elite hee a 16 Supralabial sige ek eat ein, Palen ae ome naps 6 6 6 Jintrallalbial see esc ei ees ae eee 6 6 6 Temporalseacits eons ee ee 1+2 1-2 i se72 1+3 Supralabials in eye........- sets & ae ate eo ke 3, 4 3, 4 3, 4 Ventral sii tees ey detects a oo ee oe 126 129 128 Slo caudal spe a ree sles ere nye ne Pare ? 29 28 Meaixallanyatecthisniateric meee tee cies en 19 19 20 Scalepiormulanes Aor Genes soa 15-15-15 | 15-15-15 | 15-15-15 All from Doi Suthep, Chiang Mai. Variation: The subcaudals are normally divided but occasionally some may be single. The ventrals vary from 122 to 142; subcaudals, 23 to 30. The reticulation may leave a dorsolateral series of rounded black spots connected with each other by brown cross- bars or a series of obliquely placed light brown, black-edged, cross- bars. Rarely the spots on venter may be absent. SERPENTS OF THAILAND 795 A second specimen, No. 675 9, was taken at an elevation of approximately 1,000 meters on Doi Suthep. The specimen was found crawling in short grass where it moved with considerable rapidity for a snake of its size and structure. The chevron terminates on the parietals but more or less con- nects with a blackish line which outlines the frontal and the sutures that touch the frontal. Besides the heavy black bars on venter, a great deal of the remaining surface is heavily peppered with gray pigment. Distribution: The species seems to be confined to Burma and Thailand. It has been taken at the Thai-Burmese border in the Sai Yoke district, Kanchanaburi province; at Chiang Mai, Doi Ang-Ka, and Doi Suthep, Chiang Mai province; Khun Tan. The specimen here described is from Doi Suthep. Remarks: In general appearance this snake resembles that of a species of the genus Oligodon, having a characteristic chevron on the neck. The separation of the first labials will easily separate Plagiopholis from Oligodon. There are 17 small maxillary teeth of nearly equal size, the first one or two smallest. M. Smith has united Plagiopholis Boulenger with Trirhinopholis an action followed here without certainty that this represents the true relationship. Genus Pseudorabdion Jan Pseudorabdion Jan, Arch. Zool. Anat. Fis., Genova, vol. 2, 1862, p. 10 (type, Rabdion torquatum). Diagnosis: Small snakes; eye not covered by scutes; internasals and prefrontals present; supraocular present; no anterior temporal; five or six supralabials; scales in 15 longitudinal rows, smooth, with- out apical pits; anal single, subcaudals double. The genus is distributed in the Malay Peninsula and in the Indo- Australian Archipelago as far as the Philippines. A single species occurs in Thailand. Pseudorabdion longiceps (Cantor ) Calamaria longiceps Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1947, p. 910, pl. 40, fig. 1 (type-locality “Pinang” = Penang). Oxycalamus longiceps: Giimther, The reptiles of British India, 1864, p. 199; Stoliczka, Journ. Asiat. Soc. Bengal, vol. 42, 1873, p. 120; Blanford, Proc. Zool. Soc. London, 1881, p. 218 (Singapore ). Pseudorhabdion longiceps: Boulenger, Ann. Mag. Nat. Hist., ser. 5, vol. 16, 1885, p. 389; Bourret, Les serpents de lIndo-Chine, 1936, p. 226, fig. 105; M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 57 (Bangnara = Narathiwat, 796 THe UNIVERSITY SCIENCE BULLETIN Narathiwat); Leviton and Brown, Proc. California Acad. Sci., ser. 4, vol. 29, no. 14, May 29, 1959, pp. 481-486 (literature list); Tweedie, Snakes of Malaya, 1954, pp. 51-52, fig. 12, a and b; ibid., 2nd Ed., 1957, pp. 33, 133, figs. 13, a and b. Pseudorhabdium longiceps: Boulenger, Catalogue of the snakes in the British Mus., 2nd Ed., vol. 2, 1894, p. 329; A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 154; Flower, Proc. Zool. Soc. London, 1896, p. 886; Robinson and Kloss, Journ. Federated Malay States Mus., vol. 8, 1920, p. 303; Sworder, Singapore Nat., 1923, p. 65; Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 161; Journ. Federated Malay States Mus., vol. 10, 1922, p. 267 (Frazer’s Hill, Malaya). Rhabdion torquatum Duméril, Bibron, and Duméril, Erpétologie générale A AvoleWelSp45 pp: 119- 122, (type-locality, “Macassar”); Peters, Mon. Akad. Wiss. Berlin, 1861, p. 684. Pseudorhabdion torquatum: Jan and Sordelli, Iconographie générale des ophi- diens, 1865, livr. 10, pl. 3, fig. 3. Diagnosis: A diminutive snake reaching a length of little more than 200 mm. Slate-gray above and below; scales in 15 rows throughout length of body; anal single; mental in contact with the first chinshields; internasals small, one third size of prefrontals which are but little less in size than frontal; parietal touching labials; ventrals, 141; subcaudals, 19. Description of species (from No. 35560 Bhetong, Yala province): Rostral higher than wide, part visible above forms a very small triangle; internasals less than a third as large as prefrontals, sepa- rated from labials; prefrontals very large, touching three first labials, the first very narrowly; frontal only a little larger than pre- frontals, longer than its distance from tip of snout, about three times as broad as narrow supraoculars, much shorter than parietals; nostril pierced in anterior end of single nasal; no loreal; one small preocular, one postocular, the latter touching two labials. No anterior temporal, posterior temporal very large; six supralabials, first smallest, fifth largest, third and fourth entering orbit; six infra- labials, first pair small, widely separated, first three touching first chinshields which are nearly double size of second pair and in contact with mental; eye diameter greater than distance of eye to mouth. Scale-formula: 15-15-15, scales without apical pits. Ventrals, 141; subcaudals, 19; anal, single. Color in life: Slate-gray or bluish gray, scales with slightly lighter edges; ventrals grayish brown, posterior edges slightly lighter; top of head darker than sides; some lighter marking on chin and labials. Measurements in mm.: Total length, 208 mm.; tail, 15.5; head width, 4.2; head length, 7.3; greatest width of body (female with eggs), 7mm. SERPENTS OF THAILAND 797 Distribution: The species is known in Thailand only in the two southernmost provinces, Yala and Narathiwat. Outside of Thailand the species occurs in Malaya and has been reported in Sumatra, Borneo, Celebes, and the Philippines. Leviton and Brown (1959) question the Philippine reports. Remarks: I collected this small specimen about four kilometers from Bhetong, Yala province, on the edge of a rice-paddy ditch, among dead leaves. The specimen is a gravid female containing three eggs. Another specimen was taken by me on Singapore Island. Genus Calamaria Boie Calamaria Boie, Isis, 1826, p. 981; ibid., 1827, p. 519, 539 (type of genus, linnaei); M. Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 236-237 (fixation of type). Diagnosis: Small cylindrical snakes, head not or scarcely distin- guished from neck. Eye with a round pupil, the nostril pierced in a tiny nasal; loreal absent; internasals and prefrontals fused; supra- ocular reduced; preocular present or absent; scales smooth, in 13 scalerows; subcaudals paired; no apical pits. Maxillary teeth eight to eleven, all curving, subequal. Five species of this typically Malayan genus Calamaria enter Thailand. They may be distinguished by the following key: Key To THAI SPECIES OF CALAMARIA 1. First pair of labials not in contact behind mental; chinshields and mental touch; five supralabials, third and fourth enter eye; venter ivory with numerous blackish or slate bars; tip of tail pointed. ........ .vermiformis First pair of labials in contact behind the mental; tip of tail variable... .. 2 . Five supralabials; frontal four times as wide as amcor chin and ven- ter ivory-white (head color variable); third and fourth supralabials border eve ptallepoimte de ee ease Me eee yee ee reese leucocephala IRoursstipral alo valet retin cen ee ee een eh at aay eee 3 3. No nuchal dark band bordered by yellow; nearly uniform blackish or brownish violet above; second scalerow with light dots on each scale; ven- ter ivory (pinkish in life?) ; tail ending in a rounded scale....... uniformis A nuchal dark band bordered with yellow. ..................-.2-0-5- 4 4, Five rows of black dots on dorsum; frontal twice as wide as supraocular; tail pointed, lacking paired white spots; venter white except that lateral pigment extends onto ends of ventrals.................. p. pavimentata Nuchal black band bordered fore and aft with yellow. Two (or three) pairs of yellow spots on tail; ventrals strongly edged with brown or gray, in older specimens venter nearly uniform gray............... siamensis to Calamaria vermiformis Duméril, Bibron, and Duméril Calamaria BeOS Duméril, Bibron, and Duméril, Erpétologie générale ; 7, 1854, pp. 85-87 (type-locality, Java); Jan and Sordelli, loolographic générale des ophidiens, livr. 10, pl. 2, fig. 3; Boulenger, Cata- 798 THE UNIVERSITY SCIENCE BULLETIN logue of the snakes in the British Museum, 2nd Ed., vol. 2, 1894, p. 333-335 (part.) ibid., vol. 3, 1896, p. 646; A vertebrate fauna of the Malay Peninsula; Reptilia and Batrachia, 1912, pp. 155-156 (Perak, 4,000 ft.); M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, 1916, p. 162 (“Bangnara, Patani” = Nara- thiwat, Narathiwat); Bull. Raffles Mus., no. 3, 1930, p. 57 (not recorded north of Patani = Narathiwat); de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, pp. 153-154 (part.) (Sumatra, Java, Borneo, Natuna Ids., Ternate, Malay Peninsula); Tweedie, The snakes of Malaya, 1954, p. 49, fig. 11. Diagnosis: Small snake, usually under 400 mm. in length; five supralabials, third and fourth border eye; first chinshields in con- tact with mental, separating first labials; frontal not twice as broad as supraocular; 13 scalerows. Tail terminating in a point. Dark brown or black above; head often partially or entirely yellowish or whitish; outer scalerows whitish; yellowish beneath with large black transverse spots. Description of species (from B. M. No. 1900.6.14.21 Larut Hills, Perak, 4000 feet): Rostral about one-and-one-half times wider than high, well visible above; internasals and prefrontals fused, the combined scales narrowed anteriorly, laterally touching nasals, two labials and a preocular; frontal shield-shaped about twice width of supraocular, a little shorter than its distance from tip of snout, shorter than parietals; nostril in a single small nasal scale that is about one third of area of first labial; preocular not reach- ing upper surface of head; supraocular moderately large; a single postocular touching parietal and fifth supralabial; five supralabials, fifth much the largest, third and fourth bordering eye; no anterior temporal, but one large secondary; mental as wide as lower edge of rostral, in contact with anterior chinshields, which are in con- tact with three labials, and distinctly larger than second chin- shields; the two pairs of chinshields nearly enclosing a large azygos scale that separates second pair; mental groove poorly developed. Scale-formula: 13, 13, 15. Ventrals, 164; anal, single; subcaudals (divided except last), 27; tail ending in a definite spine, directed downwards. Color: Above brownish lavender on the nine dorsal scalerows; two outer scalerows with an ivory stripe covering greater part of both rows (on outer row only on tail); infralabials, chin, and throat (for one-half inch), ivory; venter transversely barred with dark slate and ivory; bands irregular in width, and ivory bands between them also irregular and flecked with slate; lower half of supra- labials ivory, with a more conspicuous ivory-white area covering large fifth supralabial, which color is not continuous with the ven- trolateral ivory stripe. SERPENTS OF THAILAND 799 Measurements in mm.: Snout to vent, 345; tail, 43; total length, 388; width of head, 9.1; length of head, 13.5. Variation: It is a generally accepted idea that this species has what are called numerous color-forms, some very strikingly different from. the one described here. The specimen described is from a Malayan locality relatively near the border of Thailand. Distribution: A single specimen, taken in Thailand although not so stated is probably from “Bangnara” [= Narathiwat, Narathiwat] which is one of the border provinces (formerly a part of the prov- ince of Pattani). The specimens reported by Malcolm Smith in Thailand and Malaya up to 4,000 ft. all belong to this color form. Tweedie, loc. cit., however reports a form different from this as follows: “The mountain form, which occurs above 3,000 feet, is dark brown above and each scale is bordered above and below with white so that narrow white lines run along the body between the scalerows. The belly is sharply and regularly barred black and_ yellowish-white.” Tweedie gives the ventral range as 147- 210; subcaudals 15-26. Addenda: A specimen of this species acquired in 1961 at Max- wells Hill, Malaya, EHT-HMS No. 3179 ¢, has the following char- acteristics: The anterior chinshields are nearly double the size of the second pair; this second pair is widely separated by a single scale so that the mental groove is short and is very poorly indicated (as if approaching a condition in which the groove was absent; three postparietal scales (one abnormally large) between the upper secondary temporals. Ventrals, 168; subcaudals, 26; scale-formula: 13-13-13. The tail is equally as large in diameter as the body pre- ceding vent. On one side of the neck five adjoining scales on rows three and four are fused to form as many large scales (normal on the left side). Total length, 383 mm.; tail, 35mm. The two outer scalerows are cream to ivory. It would be well to reexamine the presumed color varieties to ascertain whether or not they should be regarded as subspecies. Calamaria leucocephala Duméril, Bibron, and Dumeéril Calamaria leucocephala Duméril, Bibron, and Duméril, Erpétologie générale 1854, pp. 83-85 (type-locality, unknown); Giinther, The reptiles of British India, 1864, p. 198 (“probably from British India”); Boulenger, Catalogue of the snakes in the British Museum (Natural History) 2nd Ed., vol. 2, 1894, p. 344; Flower, Proc. Zool. Soc. London, 1899, p. 674; Boulenger, A vertebrate fauna of the Malay peninsula . . . Reptilia and Batrachia, 1912, p. 157; M. Smith, Bull. Raffles Museum, no. 3, Apr. 1930, pp. 58-59. Bourret, Les serpents de l’Indochine, vol. 2, 1936, pp. 268-269; Tweedie, The snakes of Malaya, 1953, 1954, p. 50 (“pink headed reed snake” ); ? Marx and Inger, Fieldiana; Zoology, vol. 37, June 19, 1955, p. 193 (Singapore; may be a distinct local race). 800 THE UNIVERSITY SCIENCE BULLETIN Calamaria lumbricoidea var. Cantor, Journ, Asiat. Soc. Bengal, vol. 16, 1847, p. 908 (? Malaya). Calamaria_ nigro-alba Giinther, The reptiles of British India, 1864, p. 198, pl. 18, fig. c (type-locality, Penang). ? Calamaria agamensis Bleeker, Nat. Tijdschrift. Nederland-Indié, vol. 21, 1860, p. 292 (fide M. Smith) (type-locality, Sumatra ). Diagnosis: A large Calamaria reaching a length of 450 mm. The tail 48 mm. Rostral much broader than high; frontal about four times as wide as supraocular; a pre- and postocular; five supra- labials, first pair of infralabials in contact behind mental. Above blue-black, highly iridescent; chin, venter, and all or much of outer scalerow ivory white; subcaudals dimly barred with violet; a dim median subcaudal line; tail pointed posteriorly. Description of species (from No. 3178, collected at Na Pradoo, Pattani, southern Thailand): Rostral distinctly wider than high, part visible above less than half length of prefrontal suture; no internasals; prefrontals very large, a little longer than broad, broadly in contact with second labials, barely touching first labials; frontal distinctly broader than long, its width about four times as great as width of supraocular, its length equal to its distance from tip of snout, and slightly less than its distance from posterior part of parietal; parietals large, longer than wide, their length about equal to their distance from tip of snout; nostril lunate, pierced in a small single nasal that is equally as long as first labial; a preocular; supra- ocular nearly as wide as long; postocular touching two labials and parietal; no anterior temporal; secondary temporal elongate, bor- dering parietal laterally; three scales behind parietals intervening between two secondary upper temporals; a smaller median, and a pair of wider postparietals; five supralabials, second touches pre- ocular, third and fourth border eye, fifth largest, broadly in con- tact with parietal. Five infralabials; first pair broadly in contact behind mental, first three scales border first pair of chinshields; latter a little larger than second pair but scarcely as broad, second pair in contact with each other for less than one third of their length. Scales very smooth; scale-formula: 15 (occiput), 13, 18, 18. Ventrals, 149; subcaudals, 35, divided, except last which is conical; ventrals separated from chinshields by a row of three small scales. Color: Above, from snout to tip of tail, dark bluish-black, strongly iridescent. Anteriorly second scalerow ivory-white or partially so for about 60 mm.; outer row completely ivory-white anteriorly but SERPENTS OF THAILAND 801 dorsal color encroaches so that posteriorly only lower edge of row is white; fifth labial largely ivory-white, as are lower edges of other supralabials; entire chin, throat, and venter ivory-white; subcaudals with dim lavender transverse bars on each scale, and a median longitudinal lavender line of same color. Measurements in mm.: Total length, 362; tail, 54; width of head, 7.9; length of head, 9.8; width of body, 8. Variation: Considerable variation has been reported in the color of the head. Sometimes it is pure white as suggested by the name, but often this is not so. Tweedie mentions a “pink head” in some of his specimens. Malcolm Smith (loc. cit.) has shown that this variation occurs with specimens having a different marking. The preocular may be absent or present in the same series. Despite this fact, subspecific forms may exist in its relatively wide range from Thailand to Java, Sumatra and Borneo. The variation in number of ventrals for males is from 134 to 149; subcaudals, from 24 to 37. For females, 152 and 21. (Data taken by Malcolm Smith on six specimens from Sumatra. ) Distribution: In Thailand, the species is known only from Na Pradoo, Pattani, southern Thailand where a specimen was collected by Nai Pryoon Kananuracks, which he has made available to me for study. Marx and Inger (loc. cit.) refer to this form three specimens probably from Singapore Island. They report four specimens with the preocular entering the orbits which is not true of the type or specimens I have examined in the Kuala Lumpur Medical Institute or in two other specimens, one from Fraser’s Hill and the described form. It is not certain what form they had before them since they state, “These specimens agree with all published descriptions of this form.” I have not had any of the white-headed forms for examination. A Frasers Hill specimen (EHT-HMS No. 3147 3g) has 184 ventrals and 31 subcaudals. The color is similar to the described specimen, but each scale appears to have a wide transparent bor- der on the sides and tip. I have referred it to this species. Other variant characters are: the shape of the nostril is different, the nasal partially or completely divided, the second pair of chinshields narrowly in contact anteriorly, and there is a narrow contact between the first infralabials behind the mental. 31—1367 802 Tue UNIVERSITY SCIENCE BULLETIN Calamaria uniformis M. Smith (Fig. 46) Calamaria pavimentata uniformis M. Smith, Proc. Zool. Soc. London, 1921, p. 426 (type-locality, Langbian Peaks, S. Annam, 6000 ft.); The fauna of British India . . . Reptilia and Amphibia; vol. 3, Serpentes, 1943, pp. 238-239. Diagnosis: A large calamarian (455 mm.), nearly uniformly blackish to brownish violet; dorsal scales somewhat darker on edges: second scalerow with small yellowish or light violet dots on each scale; a cream-white line on lower three fourths of first scalerow; subcaudal area immaculate; scales in 13 rows; five infra- labials, the first two in contact behind mental; tail ending in a blunt rounded scale; no nuchal collar or nuchal black spot; no white spots on tail; frontal twice as wide as subrectangular supraoculars; ventrals, 192 2 , 163 ¢ ; subcaudals, 18 ¢ , 22 3. Description of species (from No. 36529, northern Chiang Mai (province), Gordon Young collector, 1960): A large calamarian; rostral as wide as high, broadly visible above; no internasals; pre- frontals almost as large in area as frontal, laterally in contact with two supralabials, a preocular and a nasal; frontal longer than wide, twice as wide as supraocular; its length greater than its distance to end of snout, equally as long as its distance to end of parietal; nostril pierced in a very small triangular nasal scale sur- rounded by three scales; one small preocular touching prefrontal, supraocular, and second supralabial; a small postocular touching parietal and two supralabials; no anterior temporal; a large elongate secondary temporal bordering parietal; no loreal; four supralabials, second and third bordering eye, fourth largest; five infralabials, first pair in contact behind mental, three anterior ones border first chinshields; second chinshields but little smaller than first pair, in contact for about three fourths of their length; vertical diameter of eye less than its distance from oral margin. Scale-formula: 15 (about occiput), 13, 13, 13; scales smooth, lacking apical pits. Ventrals, 192; anal, single; subcaudals, 18, divided (except last). Color: Above iridescent blackish or brownish violet on head, body, and tail, the periphery of scales somewhat darker; body somewhat lighter on sides; lower three fourths of first (outer) scale- row, cream or ivory; second row with lighter median spots, yellowish on anterior sixth, becoming very light violet on remainder of body and less distinct; lower part of supralabials cream-white; chin and SERPENTS OF THAILAND 803 Fic. 46. Calamaria uniformis M. Smith. No. 36529, northern Chiang Mai province, northern Thailand. Actual total length, 445 mm.; tail, 24 mm. infralabials with some dim darker clouding; throat and venter ivory to yellowish cream (pinkish in life?); subcaudal region similar to 2 venter, save that scales of lower caudal scalerow have whitish spots surrounded by darker pigment. Measurements in mm.: Total length, 455; tail, 24: width of head, 6; length of head, 11; width of body, 9. Remarks: The single specimen was collected by Gordon Young in the northern Thai province of Chiang Mai, on Doi Suthep a moun- tain arising near the city. I have examined a specimen (No. 62) presumably of this form, from Cambodia in the Vienna Natural 804 THE UNIVERSITY SCIENCE BULLETIN History Museum. The squamation of the head is practically identical. There are however, only 163 ventrals and 22 subcaudals in this male specimen. The larger number of ventral scales in the described female specimen (192, 29 more than in this specimen) may be due to sex. They are practically identical in color and markings save that the described specimen is of a darker brown and the head is a trifle narrower proportionally. In the Cambodian specimen there is a faint trace of a darker median line and the outer scalerow has the scale sutures with the ventrals pigmented which is not true of the Thai specimen. In neither specimen is there a trace of median ventral spots, and the subcaudal spots are poorly distinguished. Distribution: The species is presumably confined to Thailand, Viet Nam and Cambodia. Only known from Chiang Mai province in Thailand. Calamaria pavimentatus Duméril, Bibron, and Duméril Although several subspecies have been recognized in this species, only a single one has been found in Thailand. I regard it as the typical one. The type-locality, however, is Java. Calamaria pavimentata pavimentata Duméril, Bibron and Duméril * Calamaria pavimentata Duméril, Bibron and Duméril, Erpétologie générale ee UVLO peal appa l= 73 (type-locality, Java); Boulenger, The fauna a British India, including Ceylon and Burma; Reptilia and Batrachia, 1890, p. 282 (part.); Catalogue of the snakes in the British Museum, Ind Ed. vol. 2, 1894, p. 348 (part.); Werner, Verh. Ges. Wien, vol. 46, 1896, p. 18; Flower, Proc. Zool. Soc. London, 1896, p. 886; ibid., 1899, p. 675; Laidlaw, Proc. Zool. Soc. London 1901, p. 577; Wall, Proc. Zool. Soc. London, pp. 93, 101; Mocquard, Rev. Coloniale, 1907, p. 45; Boulenger, The vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 157; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 3, 1914, p. 213, Proc. Zool: Soc., London, 1921, p. 426; Journ. Federated Malay States Mus., vol 10, 1922, p. 267; Robinson and Kloss, Journ. Federated Malay States Museum, vol. 8, 1923, p. 364; Werner, Zool. Jahrb. vol. 57, 1928, p. 174; Brongersma, Treubia, 1929, p. 67; Smith, Bull, Raffles Mus, no. 3, Apr. 1980, p. 59. Calamaria pavimentata pavimentata ( part.) Bourret, Les serpents de l’Indo- chine, vol. 2, 1936, pp. 270-271, fig. 107. (The drawing shows the mental touching the chinshields. Two subspecies are named. ) Diagnosis: A black collar preceded and followed by yellow; five deep-black rows of dots on dorsum; first pair of infralabials in con- tact behind the mental; venter yellow; a median line under tail; four supralabials. * It is entirely possible that certain of these references actually refer to other species or subspecies. SERPENTS OF THAILAND 805 Description of species (from No. 3165, (102) Na Pradoo, Pattani, southern Thailand): Rostral about as broad as high, well visible above; no internasals; prefrontals large, touching two labials later- ally, and notched anteriorly by a minute nasal; frontal longer than wide, its sutures with supraoculars parallel, its length greater than its distance to end of snout, and greater than length of the common parietal suture; length of parietal about equal to its distance from tip of snout; nostril pierced in a minute nasal; a small preocular; supraocular less than half width of frontal; a postocular present; an upper secondary temporal borders parietal; no anterior temporal; four supralabials, second and third entering orbit, fourth very large, bordering parietal; mental broader than long; first infrala- bials in contact behind mental; five infralabials, first three in contact with first pair of chinshields; second pair of chinshields a little smaller than first pair, and in contact only anteriorly. Tail pointed. Scale formula: 17 (occiput), 13, 13, 13. Scales smooth; ventrals, 153 g ; anal entire; subcaudals 15. Color: Above rather reddish brown on body with a median and two lateral series of tiny black spots, one spot to each scale in third, fifth, and median scalerows; first, second, fourth and sixth rows with light spots, whitish on two anterior rows, but cream or very light brown on dorsum; spots larger and more regular anteriorly; top of head dark; a black band or bar on neck followed by a yellow ring; chin and throat white, the light color reaching up almost to parietal and covering lower parts of three posterior labials; venter ivory- white except that lateral color extends onto ends of ventrals and subcaudals, lightly so anteriorly, strongly so posteriorly. An in- distinct median line extends for some distance under tail. Measurements in mm.: Total length, 161.5; tail, 10.5; head, width, 3.8; length of head, 5.4; width of body, 4. Remarks: There is no absolute certainty that the form described here is subspecifically identical with the Calamaria pavimentata of Dumeéril, Bibron, and Duméril from Java. I do not regard it as be- ing subspecifically related to siamensis. The strong differences in color, in pattern, in the number of ventrals, in body proportion all suggest strongly a specific relationship. In siamensis the width of the body in total length is about 56-58 times; in pavimentata pavimentata here described the width in body length is about 40 times. S06 THE UNIVERSITY SCIENCE BULLETIN A recently acquired young specimen of Calamaria pavimentatus measures 336 mm.; tail, 25. It agrees with the described form in essential details. The ventrals are 164, subcaudals, 21. The colora- tion is brownish with a dim median row of darker flecks, broken at irregular intervals, evident throughout the body but stronger anteriorly. The second scalerow has a series of ivory spots sur- rounded by body color. The first scalerow is ivory except on upper edges of scales; some brownish flecks on last three infralabials, and a few subcaudal flecks near end of tail. Calamaria siamensis Ginther (Fig. 47) Calamaria siamensis Ginther, The reptiles of British India, 1864, p. 196 (type- locality, South Laos); Tirant, Notes sur les reptiles et les batraciens de la Cochinchine et du Cambodge, 1885, pp. 60, 66; Theobald, Catalogue of the reptiles of British India, 1876, p. 140; Boettger, Ber. Offenb. Ver. Nat., 1885, p. 121. Calamaria pavimentatus (part.): Boulenger, The fauna of British India, in- cluding Ceylon and Burma; Reptilia and Batrachia, 1890, p. 282; Catalogue of the snakes in the British Museum (Natural History), vol. 2, 1894, p. 348-349. Oe pavimentata: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, no. 3, p. 213. Diagnosis: Small, slender snakes usually under 350 mm. in length; 13 scalerows; internasals fused with prefrontals; nasal very small, no loreal; frontal longer than wide; no anterior temporal; large sec- ondary temporal; four supralabials; a black nuchal band bordered with yellow fore and aft; two or three pairs of yellow spots on tail. Ventrals all strongly edged with black, brown or gray; in older specimens most of each ventral scale covered with pigment. Description of species (from no. 36530, northern Chiang Mai, Thailand): Rostral distinctly wider than high; length of part visible above, about .4 of length of suture between prefrontals; latter scales larger than frontal, touching two supralabials laterally; frontal longer and about twice as wide as supraoculars, longer than its distance from tip of snout, more than one third shorter than parietals; nasal minute, between three scales; a small preocular; a postocular touch- ing four scales; no anterior temporals; a moderately large secondary temporal; four supralabials, fourth nearly quadrangular bordering parietal, very much larger than second; second and third border orbit; vertical diameter of eye equal to or slightly greater than its distance from oral border; five infralabials, first three touching first chinshields which are truncate anteriorly and larger than second SERPENTS OF THAILAND 807 Fic. 47. Calamaria siamensis Giinther. No. 36530, Doi Suthep, Chiang Mai, Chiang Mai province. Actual total length, 237 mm.; tail 9.5 mm. pair; latter scales in contact for a very short distance, then separated by a scale. Scale formula: 17 (occiput), 13, 13, 13. Scales smooth, lacking apical pits. Ventrals, 173; anal, single; subcaudals, 12, divided. Color: Above dark brown, edges of scales darker than their cen- ters; scale centers under magnification show numerous vermiculate white and brown marks; top of head dark blackish-olive; some 808 THE UNIVERSITY SCIENCE BULLETIN cream-white marks on lower parts of supralabials; a deep black nuchal band with a yellowish mark placed both in front and be- hind it. Much dark pigment on infralabials with some cream-white marks; distinct yellow areas on sides of head. Ventrals ivory all with some brown pigment forming a bar on each scale, pigment stronger posteriorly; subcaudals similarly marked. Measurements in mm.: Total length, 237; tail, 9.5; head-width, 4.1; head-length, 5.5; width of body, 4.2. Remarks: It has been necessary to exhume this species name for this distinctive species occurring in the mountains of northern Thailand. Measurements in mm. and Scale-counts of Calamaria siamensis NUMBER 36530 36531 3176 Go. Callies toes Nese etc tee eeemeiea Chiang Mai | Chiang Mai | Chiang Mai Se ig ae estes aT EE ee Ee fe) Q of Mo talWlen gts se es cee meee 237 248 225 Ta Pe eaten nile vel ana he ealleann 9.5 9.3 9.9 lead =widit essere ys crease mate 4.1 4.1 4.1 Headdlenctiniaes) ewer ye eer 6.3 6.3 6.3 Body=widitthiy cers ae 4.1 4 4 Rreocularniie ecu aes 1 1 il Rostocularstestm sce see 1 1 1 Supralaloralsyee sie emer rec: eles) seeder 4 4 4 nono OMNI es eos leas Gow eie eos e 5 5 5 Vientrals Sic tng es nen reine eee 173 172 172 Subcaudal Sapa eee Ww 11 13 In color and pattern the three are practically identical. The yellow band following the black nuchal band is less distinct in No. 36531. The preocular in all is narrow, and diagonally placed. There are three scales between the second chinshields and the first ven- tral; and three postparietals between the temporals. Genus Ahaetulla Link Ahaetulla Link, Beschr. Nat. Samml. Rostock, 1807, p. 73 (type of genus fasciata = Coluber ahaetulla Linnaeus (part.). Dendrophis Fitzinger, Neue Class. Rept., 1826, pp. 29, 30, Isis, 1827, p. 519, 520 (type, Coluber ahaetulla). Dendrelaphis Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 339 (type, caudolineatus ). Tachyophis Mertens, Arch. Naturg. Berlin, vol. 3, 2, 1934, p. 189 (type Coluber pictus; name preoccupied, Rochebrune 1884). Diagnosis: Maxillary teeth, 20-34, posterior ones smaller or scarcely larger than preceding teeth; head distinct, loreal region SERPENTS OF THAILAND 809 concave; scales smooth in 13 or 15 rows, outer row widened, others narrow, with apical pits, the transverse rows disposed obliquely; vertebral scales enlarged; ventrals with lateral keels on venter, the ventrals notched. Arboreal serpents. There are five species of this genus in Thailand: A. ahaetulla, cyanochloris, subocularis, formosa, and caudolineata. They may be distinguished by the following key: Key TO THE THAI SPECIES OF AHAETULLA 1. Scales in 13 rows; black longitudinal lines on tail.......... .caudolineata SCalesaia VOuRO WSs eer hee ee een ei ee eee Nate ee a, 2 2. A single elongate supralabial forming lower border of orbit; vertebrals feebly enlarged, narrower than scales of outer row; seven (or more rarely elght) SUPEAOCUIATS 5 er eee We Fee a ecient subocularis IMoresthanvaysingle labialzentering:onbltjace.) 2250 ee Gece ae ee 3 3. Three or four, rarely two, postoculars present; eye large, as long as its distance from rostral; frontal as long as parietals; nine supralabials, three lalbralssborderonbiteesas js oer ree eae ee rere mete formosa Two postoculars; three labials border orbit; eye as long as its distance fromenostiuleavertebrals-vanriablenc. se ret yaM a gerne oc a ae 4 4. No stripes on flanks; vertebrals strongly enlarged, larger than outer scale- row; a broad temporal stripe extending on to neck and body. . .cyanochloris A yellow or cream stripe on first two scalerows, bordered above and below by a black line; vertebral scales not broader than outer scalerow, ahaetulla ahaetulla These snakes all have a general similarity in appearance and all are excellent climbers. The nomenclature of this genus has been in dispute for some time. I have accepted the opinion of Dr. J. Oliver* and Dr. Malcolm Smith that the name Ahaetulla applies to these Asiatic snakes, rather than to a South American species. Ahaetulla caudolineata Gray (Fig. 48) Ahaetulla caudolineata Gray, Illustrations of Indian zoology, vol. 2, 1834, pl. 81 (type-locality not given). Leptophis caudolineatus: Cantor, Journ. Asiat. Soc. Bengal, vol. 16, 1847, p. 982. Dendrophis caudolineata: Giinther, The reptiles of British India, 1864, p. 297. Dendrelaphis caudolineatus: Boulenger, The fauna of British India, Ceylon and Burma . . . Reptilia and Batrachia, 1890, pp. 339-340; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 2, 1894, pp. 89-90; Boulenger, A vertebrate fauna of the Maiay Peninsula . . . Reptilia and Batrachia, 1912, p. 147; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, 110,05) LOIS: pp. 213:2 Jour Nat. Hust: Soc;ssiamy vol.) 2; 191165 p. 161 (“Patani”); de Rooij, The reptiles of the Indo-Australian Archipelago, vol. * Oliver, Bull. Amer. Mus. Nat. Hist., vol. 92, art. 4, 1948, pp. 167-190, 267-272; Smith, The fauna of British India . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, p. 241. Contrary opinion, Dr. Jay Savage, Bull. Chicago Acad. Sci., vol. 9, 1952, pp. 203-216. 810 THE UNIVERSITY SCIENCE BULLETIN Fic. 48. Ahaetulla caudolineata Gray. No. 34789, Na Pradoo, Pattani province, southern Thailand. Actual total length, 1183 mm. 2, Ophidia, 1917, pp. 68-69, fig. 36; Sworder, Singapore Naturalist, no. 2, 1922, p. 64; M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 52 (“Klong Wan, Tapli, Isthmus of Kra”). Ahaetulla caudolineata: M. Smith, The fauna of British India . . . Rep- tilia and Amphibia, vol. 3, Serpentes, 1943, p. 250; Tweedie, The snakes of Malaya, 1954, p. 55. Diagnosis: Anterior maxillary and mandibular teeth longest. Head distinct from neck; scales in 13 rows disposed obliquely, with apical pits; scales except median and outer rows, elongate, narrow; ventrals with lateral keels. Description of species (from No. 3478): Rostral once and a half wider than high, very narrowly visible above; internasals narrowed a little in front, about as wide as long; prefrontals broader than long; frontal longer than wide, the sides emarginate, as long as SERPENTS OF THAILAND 811 its distance from tip of snout, shorter than parietals; nasal com- pletely divided; loreal two to three times as long as high; one preocular; supraocular nearly as wide as frontal; two postoculars; temporals, 2 + 2 -+ 2, anterior ones small, posterior largest; supra- labials, 9-9, fifth and sixth bordering orbit; infralabials, 11-11, five anterior bordering first chinshields, which are wider but shorter than second pair; latter separated by scales. Scalerow formula: 19 (occiput), 13, 11; scales smooth, anterior body scales with single apical pits; ventrals, 177, keeled and notched laterally; anal, divided; subcaudals, 103. Color: Top of head brown-olive nearly uniform; supralabials greenish yellow with a little black coloration on posterior edges of third and fourth labials; chin and beginning of throat yellowish ivory; anteriorly greenish or greenish gray gradually developing green lines separated by coal-black lines bordering edges of dorsal and lateral scalerows; posteriorly the green gradually turns to greenish or yellowish ivory; a wider green stripe on edges of first and second scalerows bordered above and below by a black line, lower one near twice width of black upper border; five black lines and four yellowish lines on tail. Tail tapering gradually to a fine point. Venter nearly uniform grayish green; posterior edges of ventrals and subcaudals with an indistinct whitish line. A black median subcaudal line. Measurements in mm. (Nos. 3478, 3917, Na Pradoo, Pattani) : Total length, 1183, 1029; tail, 315, 265; width of head, 20, 18; length of head, 33, 27. Variation: Ventrals vary between 171 and 188; subcaudals, 100 and 112. Boulenger reports a specimen measuring 1520 mm., the tail being 380 mm. Distribution: In Thailand the species is known only in the peninsula. Specimens have been taken at Na Pradoo, Pattani, Narathiwat, and at Tapli near the Isthmus of Kra. It is a common snake in Southern India, Southern Burma, Malaya and the Indo- Australian Archipelago. Remarks: The described specimen is from Na Pradoo, Pattani, collected by Nai Prayoon Kananuracks. Ahaetulla subocularis (Boulenger ) (Fig. 49) Dendrophis subocularis Boulenger, Ann. Mus. Civ. Sto. Nat., Genova, ser. 2, vol. 6, 1888, Aug. 22, p. 600, pl. 6, fig. 2 (type-locality, Bhamo, Upper Burma); Fauna of British India, Ceylon and Burma; Reptilia and Batrachia, 1890, p. 338. §12 THe UNIVERSITY SCIENCE BULLETIN Dendrelaphis subocularis: Boulenger, Catalogue of the snakes in the British Museum, vol. 2, 1894, p. 89; Smith, Journ. Bombay Nat. Hist. Soc., vol. 23, 1915, p. 785; Proc. Zool. Soe. London, 1921, p. 426; Gyldenstolpe, Kungl. Sv. Vet. Akad. Stockholm, vol. 55, 1916, (3) p. 15; Wall, Rec. Ind. Mus., vol. 22, 1921, p. 159; Journ. Bombay Nat. Hist. Soc., vol. 29, 1923, p. 625; ibid., vol. 30, 1925, p. 813. Dendrophis tristis subocularis: Meise and Henning, Zool. Anz., Leipzig, vol. 99, 1932, p. 292. Ahaetulla subocularis: Smith, The Fauna of British India, Ceylon, and Burma; Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, pp. 249-250. Diagnosis: Related to Ahaetulla ahaetulla but with the last three or four posterior maxillary teeth shorter and slenderer; no black stripe along edge of ventrals; anterior median series of scales with a row of white dots tending to form a continuous line; a yellow lateral line; vertebral scales scarcely enlarged; scalerows, 15; a single much elongated supralabial under eye. Description of species (from No. 34972): Rostral wider than high, narrowly visible above; internasals about as long as broad, a little shorter than prefrontals but much smaller, suture between them about as long as that between prefrontals; latter much wider than long, touching loreal, and posterior part of nasal, laterally; frontal longer than its distance to tip of snout, shorter than parietals, its anterior width greater than that of supraocular; parietals touch only upper larger postocular; row of scales following parietals some- what enlarged; nostril large, between two nasals; loreal elongate touching two labials; supralabials, 7-9 (on right side); five scales precede subocular on left side; only number three, the subocular, touches the two following supralabials; the one following subocular is excluded from temporals; one preocular, two postoculars; tem- 1 porals, 2+ —, 2 + 2+ 2; mental much smaller than rostral; 10-11 9 infralabials, ‘sixth much enlarged, five touching first chinshields, which are larger than second pair. Scale-formula: 19, 15, 15, 11, median only slightly larger than adjoining rows and definitely smaller than outer row; lateral scales in somewhat oblique transverse rows; scales with single apical pits. Ventrals, 165, each with two keels; subcaudals, 101; anal, divided; 23 maxillary teeth, posterior ones smaller and slenderer than anterior. Color in life: Head and anterior part of neck dark olive, gradually becoming bronze-brown, which color continues to end of tail, and laterally to middle of second scalerow. A cream or fawn stripe on edges of first and second scalerows below which is a very light bronzy line occupying outer half of row one and and edges of SERPENTS OF THAILAND Fic. 49. Ahaetulla subocularis (Boulenger). No. 36140, Kaeng Pang Tao, Chiang Mai province. Actual total length, 880 mm.; tail, 234 mm. §14 THE UNIVERSITY SCIENCE BULLETIN ventrals; concealed parts of lateral scales light blue with blackish blotches at each end of blue area; these black spots are not entirely concealed so that diagonal scalerows are edged with a fine den- ticulate blackish line, less distinct or absent on latter half of body; an indistinct dark line behind eye quickly disappearing on neck; low on sides of neck the light stripe is bluish or dirty white; chin with a faint lavender wash; supra- and infralabials cream, the anterior ones with small black marks on sutures; chin, cream-white with a faint lavender wash. Measurements in mm.: Length snout to vent, 454; tail, 175; total length 629; head-length, 18.2; head-width, 10. Variation: Smith records the variation in ventrals as from 153 to 175; subcaudals, 85 to 105. Remarks: The specimen when captured was crawling on the ground in cut over forest land, the new forest trees about 10-15 ft. high, were scattered. The figured specimen is from northern Chiang Mai province. Ahaetulla formosa (Boie) Dendrophis formosus Boie, Isis, 1827, p. 542, (type-locality Java); Flower, Proc. Zool. Soc. London, 1896, pp. 883-884; ibid., 1899, p. 670; Boulenger, Catalogue of the snakes in the British Museum, 2nd Ed., vol. 2, 1894, pp. 84-85: A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 146; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, p. 64. Ahaetulla formosa: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 52 (part.); Sworder, Singapore Nat., no. 2, 1922, p. 64; Tweedie, The snakes of Malaya, 1954, pp. 54-55, pl. 3, upper fig. Diagnosis: Eye large, equal to its distance from front of nostril; internasals a little longer than prefrontals; frontal as long as parie- tals: one preocular, two to four, usually three postoculars; temporals, 2 + 2; usually three supralabials bordering orbit; scales in 15 rows; vertebrals considerably larger than scales of outer row; olive, brown, or bronze-brown with reddish and greenish shades; scales edged with black; a black stripe on sides of head, passing through eye, extending on to neck where it widens and converges toward its fellow; two black lines may be present on posterior part of body; venter greenish. Description of species (from Boulenger, 1912). Eye very large; rostral twice as broad as deep, well visible above; internasals smaller but as long as prefrontals; frontal as long as its distance from end of snout and as long as parietals; loreal much longer than high; SERPENTS OF THAILAND 815 one preocular, and two to four (usually three) postoculars; tem- porals, 2+ 2; normally nine, rarely eight supralabials, fifth and sixth enter orbit (rarely fourth also); five infralabials touching anterior chinshields which are much shorter than posterior; scales in 15 rows, those of vertebral row considerably larger than those of outer row; ventrals, 174-205; anal, divided; subcaudals, 132-158. Color: Olive, bronze-brown, or yellowish brown above, with red and green shades; scales edged with black; neck, and sometimes head, red-brown; a black stripe on each side of head passing through eye and extending onto nape, where it widens consider- ably and converges toward mid-line of neck; two black lines may be present along each side of posterior part of body; upper lip greenish yellow; lower parts greenish. Measurements in mm.: Total length (largest specimen), 1420; tail, 480. Distribution: In Thailand this species is confined to the southern peninsular area, specimens having been taken in the provinces Narathiwat, and Yala. Elsewhere the species is present throughout Malaya, and in the islands of Borneo, Sumatra, and Java. Ahaetulla cyanochloris (Wall) Dendrophis pictus var. cyanochloris Wall, Rec. Ind. Mus., vol. 22, p. 155 (type-locality, Mergui, Tenasserim, Burma). ?Dendrophis formosus: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, no. 2, Dec. 1916, p. 161 (see M. Smith, 1943, p. 244-245), Ahaetulla cyanochloris: M. Smith, Rec. Ind. Mus., vol. 42, 1940, p. 482. The fauna of British India . . . Reptilia and Amphibia, vol. 3, serpentes, Dec. 1943, pp. 244-245. Ahaetulla formosa: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 52 (part.). (“Bangnara” = [Narathiwat] and Benang Stah). Dendrophis pictus: Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, p. 337 (part.); Catalogue of the snakes of the British Museum, 2nd Ed., vol. 2, 1894, pp. 78-79 (part. fide M. Smith). Diagnosis: A slender arboreal snake, feeding chiefly on the ground. Maxillary teeth, 21-24, teeth increasing in size posteriorly; internasals shorter, or nearly as long as prefrontals; three supra- labials bordering orbit, fourth very slightly so or just failing to touch edge of orbit; vertebrals at middle of body wider than scales of outer scalerow. Scales in 15 rows reducing to nine posteriorly. Two anals. A broad temporal stripe extending onto neck where it may break up into spots. Description of species (data from Wall (1921) and M. Smith 1943): Head moderately separated from neck. Snout a little S16 THE UNIVERSITY SCIENCE BULLETIN broader and more truncate than in ahaetulla; eye as long as its distance from middle or front end of nostril; internasals as long as or a little longer than prefrontals; temporals, 1 + 1, or 2 + 2; nor- mally nine, rarely eight or ten supralabials, fourth touching edge of orbit, fifth and sixth bordering orbit; vertebral scales strongly enlarged and at middle broader than scales of outer row, their posterior margins truncate or concave. Scale-formula: 15, 15,11 (or 9). Ventrals, 186-211; subcaudals, 135-159; anal divided. Color in life: Bronze-olive above, scales black-edged; ventrals and outer scalerows pale greenish or yellowish; usually no black stripe on flank; broad black temporal stripe extending onto neck and forepart of body, where it may break into spots; lips and lower jaws yellowish. Measurements in mm.: (largest), 1330; tail, 405. Distribution: In Thailand the species occurs in the northwestern part. Outside Thailand it ranges in Bengal, Assam, Upper Burma, Tenasserim, Indo-China and the Andaman and Nicobar Islands. Remarks: This species, said to very similar in appearance to ahaetulla ahaetulla, has fewer maxillary teeth; there is a slight difference in squamation, the width of the middorsal row is greater, the details of the lateral markings are different, and the internasals are longer than the prefrontals. I have not seen a Thai specimen of this species. Ahaetulla ahaetulla (Linnaeus ) Three subspecies of this widespread form are recognized, but as yet only a single form has been taken in Thailand. Ahaetulla ahaetulla ahaetulla (Linnaeus ) (Fig. 50) Coluber ahaetulla Linnaeus, Systema naturae, 10 Ed., 1758, p. 225 (part.) (type-locality, unknown); Andersson, Bihang Till K. Sven. Vet. Akad. Stock- holm, Band. 24, Afd. 4, 1899, p. 22. Coluber pictus Gmelin, Systema naturae, Ed. 12, vol. 1, 1789, p. 1116 (type- locality, not given). Dendrophis pictus: Boie, Isis, 1827, p. 530 (Java); Boulenger, The fauna of British India . . . Reptilia and Batrachia, 1890, pp. 337-338; Catalogue of the snakes in the British Museum, 2nd Ed., vol. 2, 1894, pp. 78-79 (part.); de Rooij, The snakes of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, pp. 58-60, fig. 34; Wall, Journ. Bombay Nat. Hist. Soc., vol. 18, 1907; p. 189; ibid., vol. 19, 1909-1910, pp. 847, 788; ibid., 25, 1918, p. 509; Rec. Ind. Mus., vol. 22, 1921, p. 1538; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 1, 1914-1915, p. 86, 176, 187, 218; Sworder, Singapore Nat. no. 2, 1922, p. 64; M. Smith, Journ. Federated States Mus., vol. 10, SERPENTS OF THAILAND 817 1922, p. 266; Bull. Raffles Mus., no. 8, 1930, p. 51-52; Meise and Hennig, Zool. Anz., Band, 99, Heft, 11/12, 1932, p. 287; Shaw, Shebbeare and Baxter, Journ. Bengal Nat. Hist. Soc., vol. 14, 1940, p. 108. Ahaetulla fasciata Link, Beschr. Nat. Samml. Rostock, 1807, p. 74 (based on figure in Bechstein ). Coluber decorus Shaw, General zoology, vol. 3, 1802, p. 538, (type-locality not known). Ahaetulla bellii Gray, Ulustrations of Indian Zoology, vol. 2, 1834, pl. 80, fig. 2 (type-locality, Singapore ). Dendrophis picta andamanensis Anderson, Proc. Zool Soc. London, 1871, p. 184 (type-locality, Andaman Islands). Dendrophis proarchus Wall, Journ. Bombay Nat. Hist. Soc. vol. 19, 1909-10, pp. 347, 827 fig. (type-locality, Dibrugarh, Assam. ). Ahaetulla boiga: Cochran, Proc. U. S. Nat. Mus., vol. 77, pt. 2, p. 26; (no such species described by Lacépéde [see Pope, next entry also], Smith, 1943, p. 241). Dendrophis boiga: Pope, The reptiles of China . . . 1935, pp. 279-281. Dendrophis pictus ngansonensis Bourret, Bull. Gen. Instr. Pub. Hanoi, May, 1935, p. 4 (type-locality Ngan-son, Tong-King); Les serpents de I’Indo- Chine, vol. 2, 1936, p. 221. Ahaetulla picta: M. Smith, Bull. Raffles Mus., no. 3, 1930, p. 51. Ahaetulla ahaetulla ahaetulla: M. Smith, The fauna of British Indian Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 249-244, figs. 10A., 85. Taylor and Elbel Univ. Kansas Sci. Bull., vol. 38, pt. 2, Mar. 20, 1958, p. 1142 (Nakhon Sawan, Thailand). Ahaetulla ahaetulla: Tweedie, The snakes of Malaya, 1954, p. 54 text fig. 1. h. Diagnosis: Maxillary teeth, 23-28 increasing in length posteriorly. Eye as long as its distance from nostril; internasals shorter than pre- frontals; fourth, fifth, and sixth supralabials touch or border orbit; vertebral scales at mid-body not wider than scales of outer row; scales in 15 rows reducing to nine or eleven rows just before vent; a ventrolateral cream or yellowish line bordered by black, which may extend entire length of body. Description of species (from No. 2, Doi Suthep, Chiang Mai, circa 100 m. elev.): Head slightly distinct from neck; rostral one-and- one-fifth times as wide as high, well visible above; prefrontals large, a little wider than long, smaller and not as long as prefrontals; latter scales one-and-one-half times as wide as long, their suture with loreal and preocular subequal; frontal somewhat bell-shaped, wider than supraoculars, touching preoculars, longer than its distance to end of snout, slightly shorter than parietals; nasal single; loreal more than twice as long as wide; preocular single; two postoculars (three on right, one probably a segment from the fifth labial); temporals, 1+1+2 (2+1-42); supralabials, 8-9 third, fourth, and fifth (fourth, fifth, and sixth) bordering orbit; eye large, its length equal to its distance from front of nostril; infralabials, 10-10, five touching anterior chinshields which are shorter than posterior; latter in con- tact with one another for half their length. THe UNIVERSITY SCIENCE BULLETIN Fic. 50. Ahaetulla ahaetulla ahaetulla (Linnaeus). No. 34055, base of Doi Suthep, 5 km. west of Chiang Mai, Chiang Mai province. Actual total length, 781 mm.; tail, 216 mm. (tip missing). SERPENTS OF THAILAND 819 Scalerows: 22 (occiput), 15, 15, 9; scales rather strongly striate, with a single apical pit. Ventrals, 175, keeled on each side and notched at keel; two anals; subcaudals, 143, last three or four like- wise keeled and notched. There are 26-25 maxillary teeth, becoming considerably enlarged posteriorly. Color: Olive-bronze with a black stripe from eye, widening on occiput and passing back along sides of neck where it tends to break up and disappear. Chin and labials cream-white; top of head a lighter bronze-brown than body; a somewhat darker line on sides of snout; sides of body with a diagonal series of dark marks alter- nating with similar series of scales each with a bluish white mark, most evident on anterior third of body, tending to disappear pos- teriorly; a light stripe low on sides bordered above and below by a dark stripe. (In preserved specimens the light stripe may be partially obscured, and posterior edges of ventrals appear white ). Measurements in mm.: Snout to vent, 463; tail, 243; total length, 706; width of head 9.2; length of head 18. Variation: Two other specimens from Thailand are at hand: No. 689 ¢ Khao Chong, Trang, has 28-29 maxillary teeth; scales reducing to eleven rows posteriorly. Ventrals, 172, subcaudals, 127 +; anals, 2. No. 688 @ has 25-25 maxillary teeth; ventrals, 180; subcaudals 138; anals, 2. Locality probably Bangkok. This specimen contains six eggs. The scales bear numerous fine striations over most of their surface. (The character of the striae vary, since on the northern specimen the striae are coarser and somewhat fewer in number). The width of the median row of scales varies somewhat in the same specimen. The color characters are important but these tend to disappear in preservatives. This is not an uncommon snake, however its speed of movement in bushes and trees makes it somewhat difficult to capture. It feeds almost wholly on frogs which it obtains for the most part on the ground. Then it seemingly again returns to shrubs and trees for hiding. Most examples are encountered near water, and especially near rice fields. The species is oviparous, three to six eggs being laid at one time. Smith (1943) remarks that development of the embryo may begin before deposition. De Rooij (1917) states that they are ovovivip- arous. This, probably, is a lapsus for oviparous. Distribution: The species occurs over Thailand in suitable locali- ties and probably could be found in every province. It is also §20 THE UNIVERSITY SCIENCE BULLETIN present in India, Burma, Indo-China, Malaya, the Indo-Australian Archipelago, as far east as the Philippines. Subfamily NATRICINAE Genus Natrix Laurenti Natrix Laurenti, Synopsis reptilium, 1768, p. 73 (type of genus, Natrix vulgaris = Coluber natrix Linnaeus). The genus Natrix has long been regarded as being something of a puzzle, largely because of the diversity of the forms that have been included in it. As understood by many herpetologists it is a composite group consisting of more than a single genus. In my study of Thai species I have been convinced of the wisdom of accepting Fitzinger’s genus Rhabdophis for natricine snakes with large ungrooved back fangs formerly placed by many herpetologists in Natrix. I believe all the American species now considered in this genus are ovoviviparous. At least many of the southeastern Asiatic forms are oviparous. Malcolm Smith (1943) has declared a relationship between several small Asiatic species and Natrix modesta, synonymizing with it certain ones and suggesting that others should perhaps be so regarded. I do not regard this arrangement as acceptable; I also doubt that the group is generically separable from Natrix as has been proposed. Diagnosis: Maxillary teeth, 22-28 (piscator); 28-34; posterior teeth gradually enlarged; scalerows, 15 or 19 without nuchal groove or nuchal glands; internasals truncate (piscator) or much narrowed anteriorly. Key To Species OF Tuat Natrix 1. Nostrils lateral; anterior part of internasal a little more than half width of scale, broadly truncate; more than thirty maxillary teeth............ 2 Nostrils directed upwards; internasals narrowed anteriorly to about one- third width of the scale; less than thirty maxillary teeth............... 6 i) . Fourth, and fifth supralabials bordering orbit; usually in adults dark reticulations or dark cross-bars enclosing light spots, venter yellow, percarinata Fourth, fifth and sixth supralabials bordering orbit................... 3 3. Strong lateral dark spots or bars, usually triangular, the points usually crossing ventrals; a dorsolateral series of light spots may be present, the marks fadineswith ages 4. c8 on ee ated Ones eee trianguliygera Not: sorpmarked i. 0 ei. Sa cnet dt asi see Oe ee ee ee 4 4, Venter with three series of quadrangular dark or black spots forming — Hines se SSE GS GES SU Re ne OLN slide SO pele eR en deschauenseet Not-so marked ....3..-. 2h eee Le ee eee 5 5. Black above with a dorsolateral chain of yellow spots; two outer scalerows with yellow centers; outer edges of ventrals and subcaudals with black spots; ventrals, 147-154, subcaudals, 120-126; anal single. . . . groundwateri SERPENTS OF THAILAND 821 Indefinite gray to brown reticulation with two very narrow light dorso- lateral cream lines; labials cream, the color continued as a line diagonally on neck and joining dorsolateral lines; venter yellow, the ventrals with an OutenerayoMbplackish:spote ss aau On Se eee te inas 6. Venter yellowish or ivory; outer posterior edge of ventrals grayish or sometimes blackish; diagonal lines from eye absent or very dim; no black diagonal lateral stripe on neck tending to meet its fellow on nape, piscator Each ventral scale with a well-defined complete transverse stripe; two strong diagonal lines from eye, and a lateral nuchal black line, flavipunctata* Natrix percarinata Boulenger Tropidonotus percarinatus Boulenger, Proc. Zool. Soc. London, 1899, pp. 163-164 (type-locality, Kuatun Mts., 3000-4000 ft., NW Fokien, China). Natrix percarinata: M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 6, 1923, p. 201 (Hainan); Rec. Ind. Mus., vol. 42, 1940, p. 483; Parker, Ann. Mag. Nat. Hist., ser. 9, vol. 15, 1925, pp. 302, 304; Pope, The Reptiles of China, 1935, pp. 116-120, pl. 6, figs. F and G; Stejneger Proc. U. S. Nat. Mus., vol. 66, 1925, p. 71 (Szechwan); M. Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, 1943, p. 299. (Southern China, Upper Burma, North Siam [Doi Suthep], Hainan, Tong-King, Annam, and Formosa. ) Natrix annularis:; Barbour, Mem. Mus. Comp. Zool., Harvard Coll., vol. 40, 1912, p. 130 (Ichang) (not of Hallowell). Natrix aequifasciata: Chang, Contr. Biol. Lab. Sci. Soc. China (Zool. Ser.), vol. 8, 1932, p. 33 (not of Barbour). Diagnosis: Maxillary teeth, 30 to 34. Posterior maxillary teeth gradually enlarged; internasals narrowed anteriorly; nostrils directed slightly upwards; scales in 19 rows, keeled; two anterior temporals; two labials border eye. Description of species: Rostral twice as broad as high, barely visible above; internasals much longer than broad, narrowed ante- riorly, longer than prefrontals; frontal one-and-three-fifths times as long as broad, as long as its distance from end of snout, a little shorter than parietals. Nasal completely divided; loreal as high as wide; one preocular, three postoculars, and a small subocular; temporals, 2 + 3; eight supralabials, fourth and fifth bordering orbit; five infralabials touching anterior chinshields which are shorter than posterior. Scales in 19 rows; ventrals, 141; anal, divided; subcaudals, 71, paired. Color: Grayish olive above, sides with light-edged black vertical bars; four anterior supralabials grayish olive as is upper surface * While I am unable to accept Dr. Malcolm Smith’s arrangement of flavipunctata as a subspecies of piscator, I have followed him in using Hallowell’s name for the common variable form occuring throughout Thailand. The type of flavipunctata was a rather small specimen, something over 22 inches in length. Hallowell states: ‘“‘Head small eye some- what prominent” “eight supralabials the eye resting on the third and fourth.” He fails to describe the color and markings of the body save to state “Color dusky yellow with numerous yellow spots along the margins of the scales.” ‘Under parts yellow, the posterior margin of each scute bordered with black.” 822 THe UNIVERSITY SCIENCE BULLETIN of head. Rest of head yellowish white like lower surface; venter uniform yellowish white anteriorly, spotted and speckled with black posteriorly; subcaudal area dark gray with some black spots. Measurements in mm.: Total length, 500; tail, 130. Variation: Maxillary teeth vary between 30 and 34, the ventrals from 133 to 157; subcaudals 68 to 85. Temporals may be 3 + 3. The scales are strongly keeled, the outer row may or may not be smooth. The young are dark olive green or gray, the color de- scending on sides as v-shaped bars, often forming complete bands; area between bars yellowish. Distribution: The species has as yet been taken only in Chiang Mai province on Doi Suthep. It is widely distributed from Burma through Southern China to Hainan and Formosa, One would expect to find the species in all the northwestern provinces. Remarks: The species reaches a length of 940 mm.; the tail, 270 mm. The species is reported sometimes as brooding the eggs. Natrix trianguligera (Boie) (Fig. 51) Tropidonotus trianguligerus Boie, Isis, 1827, p. 535 (type-locality, Java); Boulenger, Catalogue of the snakes in the British Museum, 2nd Ed., vol. 1, 1893, p. 224; A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 125; Wall, Journ. Bombay Nat. Hist. Soc., vol. 31, 1926, p. 560; M. Smith, Journ. Nat. Hist. Soc. Siam, vol. 2, no. 2, Dec. 1915, p. 159 (Khoa Wang Hip, Nakhon Si Thammarat; Klong Bang Lai, “Patiyu’ ). Natrix trianguligera: Smith, Bull. Raffles Mus., no. 3, 1930, p. 43; Barbour, Mem. Mus. Comp. Zool. Harvard Coll., vol. 44, 1912, p. 108; de Rooij, The reptiles of the Indo-Australian Archipelago, vol. 2, Ophidia, 1917, p. 84; Sworder, Singapore Nat., 1922, pl. 61; Tweedie, The snakes of Malaya, 1954, p. 68, fig. 15 d. Diagnosis: Internasals narrowed and truncate anteriorly; nasal divided completely or almost so; one preocular, three postoculars; temporals, 2 + 3; nine supralabials, fourth, fifth, and sixth entering orbit; scales in 19 rows. A broad olive to olive-brown median stripe, spotted or reticulated with gray-black; sides reddish with triangular or subquadrangular spots, the apices on ventrals; venter cream (reddish in life). Description of species (from No. 20285, Boonsong Lekagul col- lection): Head distinct from neck; rostral about twice as broad as high; internasals longer than broad, narrowed anteriorly and trun- cate, the suture between them longer than that between prefrontals; latter broader than long; frontal truncate anteriorly its sides slightly concave, its length greater than its distance to tip of snout, shorter SERPENTS OF THAILAND 823 Fic. 51. Natrix trianguligera (Boie). P.K. 108, Na Pradoo, Pattani prov- ince, southern Thailand. Actual total length, 351 mm.; tail, 110 mm. than parietals; nasals seemingly completely divided, anter!or part largest; loreal approximately as high as long, or a little higher; the single preocular reaches top of head but separated from frontal; three postoculars, upper largest; temporals, 2 + 3; supralabials, 9-9, fourth, fifth, and sixth border orbit; diameter of eye approximately as long as snout; infralabials, 10-10, first five touch first chinshields which are shorter than second pair; latter separated by one or two scales. Scale formula: 21 (occiput), 19, 19, 17; scales strongly keeled except outer row which is usually smooth; ventrals, 138; anal, divided; subcaudals, 88. Maxillary teeth, 32, last three or four gradually but slightly en- larged. 824 THE UNIVERSITY SCIENCE BULLETIN Color: Head and median part of back (seven or eight scalerows ) dark-olive with a tiny pair of black spots on parietals; on back numerous black flecks or spots posteriorly becoming a dim reticula- tion; sides reddish (faded), with series of vertical black triangles, the apices on ventrals, and each spot showing one or more small light flecks; intervening ground color much wider than triangles anteriorly, but posteriorly they become equal and even smaller; triangles, as they reach ventrals, usually send out processes which may or may not cross ventrals; venter probably reddish (faded to cream); on tail lateral triangular spots almost joined together on underside of tail; labials with black sutures. Measurements in mm. (Largest known): Total length, 1200; tail, 340. Variation: The ventrals vary between 134-145, the subcaudals be- tween 86 and 96. Two young specimens are at hand. These are No. 35586, which I collected at Tonka Harbor Tin Mine, Ronpibon, Nakhon Si Tham- marat; and P.K. No. 108 from Na Pradoo, Pattani province col- lected by Nai Prayoon Kananuraks. The first specimen (35586) has 43 lateral black spots on the body, 134 ventrals, two anals and 89 subcaudals. The second has 44 lateral black spots, 136 ventrals, two anals and 95 subcaudals. Four is the usual number of postoculars, but the first specimen has only two on the right side. The lateral spots on this specimen are gen- erally triangular but in the second they are nearly quadrangular and from the lower edge of each are two narrow taillike projections that extend onto the ventrals or cross the venter. On the tail the intervening light marks: are very narrow and become obsolete to- ward the tip of the tail. The second specimen (No. 108) of Natrix trianguligerus, from Na Pradoo, varies in certain other characters from the described speci- men. The head is moderately slender; the rostral is at least 1 times wider than high and is visible above as a line. The internasals are narrowed anteriorly and are longer than wide, while the prefrontals are broader then long, their mutual suture shorter than that be- tween the internasals. The frontal is longer than wide, greater than its distance to the tip of the snout, but shorter than the parietals; the supraocular is only slightly less wide than the frontal. The nasal is almost or completely divided, the nostril chiefly in the anterior part. The loreal is higher than long, the preocular preceding it SERPENTS OF THAILAND 825 rather large, reaching the upper surface of snout. Four postoculars are present, the upper largest, touching the parietal. The temporal formula is 2-++ 2. There are 9-9 supralabials, the 4th, 5th, and 6th enter the eye, the 8th largest. The diameter of the eye is three- fourths of the length of the snout. The mental is not as wide as the rostral. The infralabials are 9-9, five touching the first chinshields, that are distinctly shorter than the second pair which touch only anteriorly. One pair of scales is present between the second pair of chinshields and the first ventral. The scale-formula: 23 (occi- put), 19, 17, 17, all rows keeled except the outer. Distribution: The species is known from Pattani peninsular Thai- land, and it has also been reported from Nakhon Si Thammarat by Malcoln Smith. The specimen here described is from near Ranong, Ranong province at the Isthmus of Kra. The species occurs also in Burma, Malaya, and the Indo-Australian Archipelago. Natrix deschauenseei Taylor (Fig. 52) Natrix deschauenseei Taylor, Proc. Acad. Nat. Sci. Philadelphia, vol. 86, 1934, pp. 300-302, plate 17, fig. 5, text figs. 2, 3, 4 (type-locality, Chiang Mai, North Thailand). Natrix modesta: Smith, The fauna of British India, Ceylon, and Burma Reptilia and Amphibia, vol. 3, Serpentes, 1943, pp. 291-292, (part.). Diagnosis: Maxillary teeth, 36, the posterior three or four en- larged but not abruptly so; eight supralabials; scales in 19 rows. Dorsal scalerows, except outer two or three, keeled. Venter with three rows of gray spots. Description of species (from No. 5, Doi Suthep, Chiang Mai, topotype): Head distinct from neck; rostral a little less than twice as wide as high, narrowly visible from above, truncate posteriorly; internasals narrowed, truncate anteriorly, their length equal or slightly less than that of prefrontals; latter about one fourth wider than long, the longest lateral suture with loreal; frontal shorter than its distance from tip of snout, distinctly shorter than parietals; latter scales distinctly shorter than wide; nostril in a single nasal that shows a partial division below nostril; loreal large, consider- ably longer than high; two preoculars, upper largest, and two postoculars, parietal touching the upper, the first temporal touch- ing both; temporals, 5 + 2+ 8; supralabials, nine, fifth and sixth border orbit; infralabials, nine, five touching anterior chinshields which are equally as large as posterior; scales keeled except on three outer rows anteriorly, and two outer rows posteriorly. §26 THe UNIVERSITY SCIENCE BULLETIN Fic. 52. Natrix deschauenseei Taylor. No. 35952, Doi Suthep, Chiang Mai province. About actual size. Scale-formula: 23 (on occiput), 21 at fourth ventral, 19 at middle of body, 17 on latter part of body; apical pits absent (at least I discern none); ventrals, 159, that preceding the divided anal scale is also divided; subcaudals, 140. Color in life: Brownish gray, anteriorly with two lighter dorso- lateral series of dots or X-shaped spots, separated by ground color, and which farther back form a more or less continuous lighter line covering parts of three scalerows on body, only one on tail; between light lines five median rows have series of very indefinite SERPENTS OF THAILAND 827 dark spots or flecks that may be dimly joined on back but not visible on tail; scales of four outer scalerows with light centers and dark edges; ventrals edged in black making a fine continuous line bor- dered by a white line. Venter with three rows of brownish gray spots, the median beginning at about eleventh ventral, lateral ones somewhat farther back. As vent is approached, spots cover most of venter; on latter half of body there is a peppering of pigment between spots; subcaudals brownish, scale-edges lighter; outer edge of subcaudals with a continuous light line. Head variegated brown with indefinite blackish flecks and lines; supralabials light with darker edges, especially the suture be- tween sixth and seventh; a dark line begins behind eye, curves down across eighth and ninth labials with a lighter line above it which joins the dorsolateral line. Chin, throat, and infralabials cream, latter scales with blackish sutures. Measurement and data on Natrix deschauenseei NUMBER 5 67 68 Rotalplent hime wea ee 480 363 290 Sileeree eee ay a eae ca NE ee 172 124 101 Tlie adBwacltlne tz. rs ee ac al Ao as 8 8 6.5 Feadslenethietia ek ss ee pees bi 14 12 11 SllToralaloialseepeaw se rae cae 9 9-9 9-10 |hounrAlAloreils.s 45 2s cca bea clbo oben bos 9-9 10-10 9-9 Labials entering eye. ...:......:.... 5, 6 5, 6 5, 6 (5, 6, 7) Rreocwlarsseciact ha 30 oe ee 2 2 2 Rostocullanses ios tee ee ee 2 2 2) Memporal stoned Wo rites eae aS 1+2+3 14+2+3 1+2+3 Wie rare all Speco mcmama eae oe pee IFO) ee Sr cen ego ne wie ete al ane eee t ree SullooennGllS= ose secu sbocc segs: AQ aes ier eee deere eal Rae ieee ce Distribution: Known from type-locality only. Natrix groundwateri M. Smith (Fig. 53) Natrix groundwateri M. Smith, Joe Nat. Hist. Soc. Siam, vol. 4, no. 4, 1921, pp. 205-206, pl. 8, fig. 2 (type-locality, Tasan 40 km. S. W. Chum- phon, PRanong province, Thailand): M. Smith, The fauna of British India : Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1948, p. 291. Diagnosis: Small slender Natrix; 19 scalerows anteriorly, reduc- ing to 17 at middle of body; scales smooth anteriorly, feebly keeled posteriorly; ventrals, 147-151; subcaudals, 120-126; supralabials, nine, the fourth, fifth, and sixth entering orbit. §28 Tue UNIVERSITY SCIENCE BULLETIN Description of species (after M. Smith): Rostral much broader than deep; internasals narrowed anteriorly, as long as prefrontals; frontal one and one-fourth times as long as broad, as long as its distance from end of snout, much shorter than parietals; loreal twice as long as high; two preoculars; two or three postoculars; temporals, 1 + 2; nine supralabials, fourth to sixth touching eye; five infralabials in contact with anterior chinshields which are as long as posterior. Fic. 53. Natrix groundwateri M. Smith. From M. Smith, Journ. Nat. Hist, vol. 4, 1929, pl. 8, fig. 2. Maxillary teeth 30, last not abruptly enlarged. Scales in 19 rows anteriorly reducing to 17 at mid-body, and so continued to vent, smooth anteriorly, feebly keeled posteriorly. Ventrals, 147; anal, single; subcaudals, 120, divided. Color: Black above with a dorsolateral chain of yellow spots connected by a festooned stripe; two outer rows of scales with yellow centers; labials yellow with black sutures, and a yellow streak from angle of mouth to nape; a v-shaped mark behind occiput. Below yellowish, with a black spot at outer edge of each ventral and subcaudal shield. Measurements in mm.: Total length, 450; tail, 165. Variation: Four other juvenile specimens taken in the same locality show practically no variation from the type-specimen except in the number of the ventral and caudal scales and in the color of the head which is light brown above, with yellow vermiculations. Ventrals vary between 147 and 151; subcaudals, 120 and 126. Distribution: Known only from the type locality. SERPENTS OF THAILAND 829 Natrix inas (Laidlaw ) (Fig. 54) Tropidonotus inas Laidlaw, Proc. Zool. Soc. London, 1901, vol. 2, p. 576, pl. 35, figs. 3 and 4 (type-locality, Gunong Inas, Perak, Malaya); Boulenger, A vertebrate fauna of the Malay Peninsula . . . Reptilia and Batrachia, 1912, p. 125; M. Smith, Journ. Nat. Hist, Soc. Siam, vol. 2, 1916, p. 159. Natrix modesta: (part.) M. Smith, The fauna of British India, Ceylon and Burma . . . Reptilia and Amphibia, vol. 3, Serpentes, Dec. 1943, pp. 290-291. Natrix inas: M. Smith, Bull. Raffles, Mus., no. 3, Apr. 1930, p. 43 (Nakhon st yam et Mts.); Tweedie, The snakes of Malaya, 1954, pp. 70-71, fig. Diagnosis: A slender snake. Ventrals, 143 to 151, subcaudals, 93 to 109. Ventrals with quadrangular dark spots on ends; ventral sur- face cream-white. Labials with a cream line turning up on sides of neck and continuing dorsolaterally as a series of small indefinite cream spots or as a dim light line. Two or three labials border orbit. Description of species (from No. 34152): Rostral stands nearly vertical, twice as wide as high, visible above as a black line: Inter- nasals about as wide as long, narrowing anteriorly; prefrontals wider than long, touching loreal, nasal, and preocular, latter forming the longest suture; frontal distinctly longer than wide, six-sided, a little wider anteriorly than posteriorly, longer than its distance from tip of snout, distinctly shorter than parietals; nasal divided, posterior part the larger, touching two supralabials; loreal nearly quad- rangular; preocular elongate reaching top of head but not touching frontal; supraocular little more than half width of frontal and some- what shorter; three postoculars upper largest; anterior temporals, two, the upper a tiny scale; temporals, 2-+ 1+ 2; a series of six scales border parietals behind, the two median scales occupy space in a deep median emargination of the parietals; parietals much narrowed posteriorly; supralabials, nine-nine, fourth, fifth, and sixth border orbit, eighth largest (as if a secondary temporal had been fused to it): infralabials ten, last smallest, first five border first chinshields which are wider but shorter than the slender, posteriorly pointed, second pair; latter separated by a row of three small scales followed by a pair of scales. Eye large, its diameter equal to its distance from nostril. Scalerow-formula: 23 (occiput), 19, 19, 17; anteriorly at least eight scalerows dimly keeled, posteriorly 15 or 17 rows keeled, the keels of median ones most distinct. Ventrals, 151; anal, divided; subcaudals, 101. ore Go S THe UNIVERSITY SCIENCE BULLETIN Fic. 54. Natrix inas (Laidlaw). B.M. No. 1916.3.27.31, Nakhon Si Tham- marat, Thailand. Actual total length, 380 mm. Color: Above olive-brown with a dorsolateral row of lighter spots anteriorly on fifth, sixth, and seventh scalerows which tend to join and form a continuous row posteriorly on the sixth scalerow; back with indefinite minute paired darker spots, often with a light fleck in center; ventral scales with outer edges dark, bordered below by a narrow cream spot and then a dark spot, the latter forming a broken line on ventrals; median two thirds of ventrals and sub- caudals immaculate cream. Head brown above with darker vermiculations and a pair of small cream spots on middle part of parietals; light cream stripes on upper lips from snout, across angles of mouth, converging, but not meet- SERPENTS OF THAILAND 831 ing, on neck, interrupted by black marks on sutures of anterior supra- labials and posteriorly bordered above and below by black. Chin immaculate except for small black marks on sutures of infralabials. Measurements in mm.: Snout to vent, 333; tail, 169; total length, 502; width of head, 8: length of head, 16.2. Variation: The type was a much smaller specimen measuring 397 mm. with the ventrals 143, the subcaudals, 96. Laidlaw states “internasal shield broadly truncate, hinder maxillary teeth gradually enlarged.” The total known variation in ventrals is 143 to 151; of the subcaudals, 96 to 109. Distribution: In Thailand the species is known from Nakhon Si Thammarat and Prachuap Khiri Kahn. Outside of Thailand the species ranges in Malaya. Remarks: The type-specimen was taken in the northern part of Malaya near the base of the mountain Gunong Inas, by Laidlaw and Yapp, at or near Sira Rimau. The specimen here described is presumed to have been taken at Huay Tam Phra, Prachuap Khiri Khan. The specimen (figured) from Nakhon Si Thammarat, (B.M. 1916.3.27.31) agrees in detail with the markings of the described specimen. In this only two supralabials enter the orbit. Natrix piscator (Schneider ) (Fig. 55) Hydrus piscator Schneider, Historiae amphibiorum naturalis et literariae, fasc. primus . . . 1799, pp. 247-258 (type-locality, Coromandel Coast, by inference, the species based on Russell’s Indian Serpents, vol. 1, pl. 33 Neeli Koea (Nihli Koeahat [Kuahat] as stated by Schneider). Tropidonotus piscator (part.) Boulenger, The Fauna of British India, Ceylon, and Burma . . . Reptilia and Batrachia, 1890, p. 349. Natrix piscator (part.) M. Smith, The fauna of British India, Ceylon, and Burma . . . Reptilia and Amphibia, 1943, pp. 293-298, figs. 95, 96 (Good synonymy and literature list). This pictured form with the brief description given by Russel is of a snake lacking two black bars extending downwards and back- wards from eye and a dark mark on the sides of the neck that is usually continuous with the second black bar; and having groups of small yellowish or cream spots on each side of the dorsum, to- gether with a large number of darker spots often quincuncially arranged, the larger ones anteriorly, the smaller and less distinct ones posteriorly. The venter and subcaudal areas are yellowish white, rather than with black bars on each ventral. §32 Tue UNIVERSITY SCIENCE BULLETIN In northern Thailand in Chiang Mai (city), Chiang Mai, there are two species of Natrix each about as common as the other, liv- ing together in the same pools. One of these I identify as Schneider's piscator. The other I regard as Amphiesma flavipunctata, which Hallowell originally described from Kwangtung province, China. It is unthinkable to regard them as subspecies of a single species. The Chiang Mai specimens were acquired by Dr. L. D. Brong- ersma of the Leyden Museum, and myself, Dec. 5 to 8, 1958, while a tiny swamp, situated some 50 meters back of the Railway Hotel, was being fished before its seasonal drying. Approximately 14 specimens were taken of the larger darker species, agreeing in most details with Schneider’s piscator; and some 30 specimens of a sec- ond species with clearly defined black neck markings, each ventral and subcaudal strongly barred with black. There was no evidence whatever that these forms were intergrading. The first species showed but little variation, while the second one displayed con- siderable variation in the dorsal pattern and the presence of red- dish coloration in some specimens. However, every specimen showed clearly the black marks on ventrals. It cannot be a sexual difference since both males and females are in both groups. That two species are involved I believe there can be no question, despite strong similarity in the structural features. In Dr. Malcoln Smith’s treatment of the species piscator he seemingly recognizes but a single form occurring in Thailand, which he calls piscator flaviven- tris (Hallowell) and it is certainly the second species mentioned above since he describes it as having the ventrals bearing black bars, and the occasional presence of reddish coloration on the body. The two may be differentiated by the following: 1. Larger snakes often reaching more than one meter in length, larger, usually quadrangular black spots arranged in longitudinal and diagonal rows, be- coming smaller and less conspicuous posteriorly; a dorsolateral series of cream or yellow-cream spots covering parts of three or four scales extend- ing more than half length of body; no distinct diagonal black mark on sides of neck; black diagonal marks from eye directed downwards and back- wards dim or absent; chin, venter, and subeaudal region yellowish, the outer edges of ventrals grayish....................... piscator piscator . Somewhat smaller snakes not reaching a meter in length; usually a row of vertical black spots on sides touching ventrals with a series of small black spots on back, separate or touching, sometimes forming a reticulum, lacking evident cream or yellow-cream dorsolateral spots but with very small dashlike yellowish marks on scale edges; two black lines from eye distinct, usually the posterior touching a diagonal black stripe on side of neck tending to join its fellow on nape; a pair of small ocelli on parietals and a light black-edged mark on occiput; ventrals each with a narrow black bar except first three or four. Occasional specimens showing con- siderable red on side of head, neck, and body............. flavipunctata bo SERPENTS OF THAILAND 833 Natrix piscator piscator (Schneider ) Certain subspecific forms of this species probably exist. I believe, however, that Natrix piscator flavipunctata of Smith (1943) should be regarded a distinct species from piscator. The two forms piscator and flavipunctata occur often in the same ponds, each maintaining its own specific characteristics without intergradation. The same is equally true of Natrix asperrimus* and of a form of flavipunctatus in Ceylon. Since there is no indication of intergradation between the two Thai forms, I propose to treat them as species as has been done by many previous workers. The very large synonymy in the literature suggests much variation in the various populations and makes evi- dent the difficulty of placing names and references without exami- nation of the specimens on which the names were based. The table shows their similarities. Diagnosis: See key above. Description of species (from M. 34947, Phu Kading, Loei Prov. ): Rostral one-and-one-third times as wide as high; internasals nar- rowed to one third of their width anteriorly, distinctly shorter than prefrontals; latter very little wider than long, suture with nasal very short; frontal longer than its distance from end of snout, one sixth shorter than combined parietals; nasal divided, anterior part much the larger, nostril near upper border of anterior part; loreal longer than high; one large preocular reaching top surface of head; three postoculars, the subocular the largest, almost reaching to below middle of pupil below eye. Temporals, 2 + 3; supralabials, 9-9, fourth and fifth bordering orbit rather narrowly; length of eye less than its distance from nostril; infralabials, 10-10, first five touch- ing first chinshields which are shorter than second pair; latter sepa- rated by small scales. Scale-formula: 24 (occiput), 19, 19, 17; farther back on neck median scales lightly keeled, five outer rows smooth (posteriorly only four or three rows smooth); scales not narrowed and elon- gated. Ventrals, 132; subcaudals, 80; anals, 2. Maxillary teeth, 23, grad- ually increasing in size posteriorly. Color: Olivaceous above with black spots quincuncially arranged, the spots in five or seven rows anteriorly but more or less con- nected by a dim reticulum; a row of cream or yellowish dots on ® See Taylor, Univ. Kansas Sci. Bull., vol. 33, pt. 2, 1950, pp. 567-659, pl. 15, fig. 2, and pl. 21, fig. 1. 32—1367 THE UNIVERSITY SCIENCE BULLETIN Fic. 55. Natrix piscator piscator (Schneider). No. 33843, Chiang Mai (city), Chiang Mai province, northern Thailand. Actual total length, 776 mm.; tail, 215 mm. SERPENTS OF THAILAND 835 each side disappearing or becoming very dim beyond middle of body; labials dirty white with considerable pigment; a pair of diagonal lines from eye extend to mouth, the posterior reaching angle of mouth; area between these lightly pigmented. Chin, infralabials, entire venter and subcaudal region, cream-white; lateral border of ventrals grayish; lateral edges of subcaudals blackish, forming fine zigzag lines. Measurements in mm.: Total length, 844; tail, 264; width of head, 22: length of head, 37. Variation: A female specimen from Phu Kading Loei province differs a little from the preceding description as follows: the pre- frontals are narrowed to about one half of their width; the anterior nasal is only slightly larger than posterior; the eighth labial is crowded out of the labial row by contact between ninth and seventh on border of mouth. There is a single elongate temporal followed by two secondary temporals. Ventrals, 144; subcaudals, incom- plete, tip of tail missing. The yellow lateral dots are often confined chiefly to interstitial skin and unless the skin is stretched, are not completely visible. The venter is immaculate save for a very narrow vague edging of gray on the outer part of ventrals. Distribution: In Thailand this species seemingly is confined to the northern part. It is very common about Chiang Mai as is another form, flavipunctata. In the small pond just back of the Railway Terminal Hotel in Chiang Mai, approximately 30 speci- mens of Natrix flavipunctata and 14 piscator were obtained by Dr. L. D. Brongersma and myself in December, 1957. Natrix flavipunctata (Hallowell) (Fig. 56) Amphiesma flavipunctatum Hallowell, Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 503, (type-locality, Canton River, Kwangtung, China). Natrix piscator flavipunctata: M. Smith, The fauna of British India, Ceylon and Burma . . ._ Reptilia and Amphibia, vol. 3, Serpentes, 1943, p. 296; Taylor and Elbel, Univ. Kansas Sci. Bull.; vol. 38, pt. 2, 1958, pp. 1151- 1152, (Khon San, Chaiyaphum province, Thailand). Diagnosis: Scales weakly keeled anteriorly on all save three smooth outer rows (more posteriorly two rows or one without keels ); posteriorly the keels are strongly developed; anteriorly inter- nasals narrowed to about one third of their greatest width; fourth and fifth labials entering eye; temporals, 2 + 3; scales in 19 rows at middle. Ventrals and subcaudals each with a distinct black 836 THE UNIVERSITY SCIENCE BULLETIN Fic. 56. Natrix flavipunctata (Hallowell). No. 33406, Chiang Mai (city), Chiang Mai province, northern Thailand. Actual total length, 701 mm.; tail, 197 mm. posterior border; series of small dashlike diagonal yellow spots present on sides. Description of species (from No. 789 Chiang Mai [city] ): Rostral twice as wide as high, narrowly visible above; internasals narrowed to one third of their width anteriorly, a little shorter than prefrontals; latter only slightly wider than long; frontal slightly SERPENTS OF THAILAND 837 widest anteriorly, sides concave, longer than its distance from end of snout, little shorter than parietals; nasal completely or almost completely divided, anterior part somewhat the larger; loreal a little longer than high, touching three anterior supralabials; one preocular barely reaching upper surface of head; three postoculars, upper as large as lower, latter not reaching down on side of head as far as middle of pupil. Two anterior temporals, upper the smaller, three secondary temporals not well aligned; supralabials eight, fourth and fifth border orbit, sixth and seventh largest; eye slightly shorter than its distance from nostril; infralabials ten, five bordering first chinshields which are shorter than second pair; latter separated by small scales. Scale-formula: 24 (occiput), 19, 19, 17, all keeled except three (two or one posteriorly) outer rows, rather weakly keeled ante- riorly, but becoming stronger posteriorly. Ventrals, 132; sub- caudals, 77; anal, divided. Color in life: Head dark olive, the labials lighter. Two black parallel lines from eye directed downward and backward; space between dark lines, yellow or yellow-olive; two black lines from mouth angles converge on neck (or narrowly fail to touch); two tiny median black parallel lines on occiput. Body olive with black spots anteriorly, distinct laterally, rather obscure dorsally, but soon all more or less commence forming a reticulum, less distinct pos- teriorly; spots on tail absent or obsolete; venter greenish white, each ventral scale with a distinct black line on its posterior edge, and many body scales with small dashlike yellow marks. Measurements in mm.: Total length, 567; tail, 164; width of head, 14; length of head, 24. The species reaches a length of 1200 mm.; the tail, 300 mm. (see table). Distribution: The species is widely distributed in Thailand. Specimens are at hand from northern Chiang Mai. Chaiyaphum, Ubon, and Phattalung provinces, and the city of Bangkok. It is one of the commonest species in the country. Outside Thailand the species is found in southern China, India, Indo-China, Assam and northern Malaya. Remarks: A large female specimen captured in the city of Bang- kok contained 60 embryos. Genus Opisthotropis Ginther Opisthotropis Giinther, Ann. Mag. Nat. Hist., ser. 4, vol. 9, p. 16 (type of genus, O. atra); Pope, The reptiles of China, Nat. Hist. Cent, Asia, vol. 10, 1985, pp. 164-166. THe UNIVERSITY SCIENCE BULLETIN 838 "WOgN “HOGA OOGSE-LESFES *190T WoIF OT-SO8TS “Te SueryD wos ore ‘QZ pur “goggs—sZEEE Yeemjoq siequINN “9}0N 19 LVS OI—OL 6—6 STG &—& Teel a hz, Gol CaS Soak a it a 006SE aneet © Ader ec eugkelliayaue ewes.