iy i" _ oe £8] . rial irc oa analy : a“ ia . 1G d ; - — S ’ oa Fae ihe at : : y i x ae i ie ba 4 Ps i a eas : ai a * past e orig - he 7 7 aieaee heey te a oe . ey Me Pe P es ns iy ‘ . : . ‘ earare a . rie = a i 7 a sg ihe 7 ; ; ‘ . sia ‘ ; a ee . - a a . te a fe hee ; ; ; 7 ey ie _# : oe ¢ ; oe ‘ ae Mi Daye tee “ pie ae eee 7 Pac wee Paved arate pokes a ee footie BIOLOGICAL BULLETIN OF THE Marine Biological Laboratory - Mee WOODS HOLL, MASS. EBditorial Stat E. G. Conxurn— Zhe University of Pennsylvania. Jacques Lors—TZhe University of Calfornia. T. H. Morcan— Columbia University. W. M. WHEELER—American Museum of Natural flistory, New York. C. O. Wuitman— The University of Chicago. EE. B. Witson—Columbia University. Managing Editor FrRaNK R. LILLIE— Zhe University of Chicago. VoLuME XIV. WOODS HOLL, MASS. DECEMBER 1907 TO MAY 1908. 202058 eta GONTENTS OF VOLUME: AIV. No. 1. DECEMBER, 1907 Moule L. On the Light Receptive Function of Ue Waren Lapile of GOnLONCMUS .... iui dieedetekencno eens: SHELFORD, V. E. Preliminary Note on the Distribution of the Tiger Beetles ( Cictndela) and tts Relation to LP VEE SOROS OS SUC: Cig kA ORE EE PEED SDN RASS SAN Ketitocc, Vernon L. Artifictal Parthenogenests in the Fe OM TCM EERE oO oy a ks seve si MDD = Sau uieraeiaa ne = BecxwituH, Cora J. Zhe Early Development of the Lat- MOTE NS SUCHE Of AMA COLVE \..) nde Gores Mass estes Reeves, Cora D. The Breeding Habits of the Rainbow Darter (Etheostoma ceruleum Cee) a Study tn OLGA SU BONO ices, Se RRO 8 ce ROR PRR No. 2. JANUARY, 1908 Rice, Epwarp L. Gill Development in Mytilus Se ate McGii1, Carouine. Zhe Affect of Low Temperatures on MarsHatt, WM. 8S. Amztosis in the Malpighian Tubules of the Walking-stick (Diapheromera femorata)......... Hareitt, Cuas. W. Notes on a Few Celenterates of Woods Floll....... Ree alo Men Dratentes asta Sia et Ge ae tee UN No. 3. FEBRUARY, 1908 ANDREWS, E. A. The Annulus of a Mexican Crayfish.... EYCLESHYMER, ALBERT C. and WiLson, JAMES MEREDITH. TEEN OCS RO EAIOD OF AMUN Sie ath Sa lees aye cise Morean, T. H. Some FPurther Records Concerning the Physiology of Regeneration in Tubularia.............6064- Rippie, Oscar. The Cause of the Production of ** Down” and other Down-like Structures in the Plumages of PRIUS 5s tt Re Se OA Ee OR EO Re MRT co ca 8b ap nn tien Pricer, JoHN Lossen. The Life Flistory of the Carpenter ALOE Ss CREEP ASB: SOTA Heat Rosa inne hk a tae No. 4. Marcu, 1908 Giaser, O.C. A Statistical Study of Mitosis and Amitto- nS) 163 177 sts tn the E-ntoderm of Fasctolaria tulipa var. distans. 219 iv CONTENTS MoopigE, Roy L. Zhe Clasping Organs of Extinct and Pecent AMPRheOrd ety wae ee eee Ret ee eaten 2 acy oan No. 5. APRIL, 1908 TANNREUTHER, Geo. W. The Development of Hydra..... Exits, Max Mares. Some (Notes on the Factors Countrol- ling the Rate of Regeneration in Tadpoles of Rana Clamata— DAUd tn: Ace Rane Ne ear see ees ee. «hace Wricut, ArperT H. Wotes on the Breeding Habits of AML OLYSCOMLG, PUMLCE A OLUTOR ent cidncm : lat eins letersi ene™ < Loc. ctt., page 300. 4 L. MURBACH. he adds (p. 302) that animals, whose tentacles were removed in such a way as to destroy the marginal bodies, never reacted to light ; this, casually mentioned, as if it were not important. Yet this should be the crucial experiment. Now since the marginal papillz in Gonionemus are not true tentacle bulbs, and are only connected with the tentacles by the outer layer of ectoderm, and since the tentacles when pulled off by merely seizing them always, in my experience, break off at their bases, leaving the latter, I could not see how the operation above referred to could be performed. It would be possible by seizing papilla and tentacle-base with forceps, but this would re- move the margin of the bell with the nerve ring. A ae PROBLEM. These considerations led me to make experiments, first: to determine whether the light reactions of Gonzonemus are dependent on stimulation of the marginal papillz, or, second: whether any other set of organs, ¢. g., the tentacles, the velum, or the gonads, also have this function. METHODS. The first experiments were made by cutting the papilla away from the margin with fine sharp scissors in such a way as to in- jure as little as possible the adjacent parts. Operated animals were left until they had recovered from the shock, or even for several days, but never long enough to allow the papillz to re- generate. They were always compared with normal animals taken at the same time. Colors were also noted, as was the tempera- ture of the water. In some of the experiments a 5 cm. deep saturated alum solution was interposed between the animals and the sun, but I have found with other experimenters, that this gives no advantage over plain water. The temperature varied no more than from one fourth to one and one half degrees centigrade in any of the experiments during an observation, and since the only purpose was to test the mere reaction of the papilla or other organs, this change in temperature could be ignored. The med- usze were exposed in dishes of white glazed earthenware, in glass dishes, and in dishes with black substratum. It may LIGHT RECEPTIVE FUNCTION OF GONIONEMUS. 5 be remarked here that in the latter case the reactions were generally slower. The exposure was made in sunlight varying as it does during the middle portion of the day. For shutting out the sun absolutely opaque material was used ; in some cases plain covers, in others covers that would completely darken the dish. While these give varying results in reaction time, they will not be considered here since constant conditions were main- tained throughout any one experiment. In accurate time reac- tion experiments there should be some means of determining the exact intensity of light used. In some cases the exposure was uniformly for 60 seconds; in the earlier experiments only until reaction took place. The rest period I varied from 30 to 300 seconds ; in general I found the former too short and the latter too long. Contraction leading to displacement was counted a reaction. From five to ten observations were made in succession and tabulated, but space will not here be taken for the tables in full, Only about one half the averages obtained will be given. It will be seen that I have followed somewhat generally the methods of Romanes,’ Conant and Berger,’ and Yerkes.? Having received response to light in all cases where the mar- ginal papillae had been carefully removed, another cause of the reaction was looked for. Now as there is a thin welt of tissue connecting the papilla which appears to be made up of similar cells, having about the same color, this welt of tissue might be conceived as taking over any light-percipient function that the removed papillae were thought to possess; it was accordingly also removed. As this is a much more painstaking operation the work necessarily progressed slowly. In some cases a method was employed somewhat similar to the one previously adopted in removing the otocyst organs in the same medusa.* In other cases I depended on cutting away as little of the margin as pos- sible and yet removing the welt of tissue in question, together with the papilla. Such operated medusz were usually tried in 1Romanes, G. J. (’76), ‘‘ Jelly-fish, Star-fish and Sea-urchin,’’ International Sc. Series, 1885. 2 Berger, E. W., Wem. Biol. Lab. Johns Hop. Univ., Vol. 1V., 4. Soom cll. 4Murbach, L. (’03), ‘‘ The Static Function in Gonzonemus,”’ Am. Jour. Physiol., Vol. X., No. 4, December, 1903. 6 L. MURBACH. from one or two to three days after the tissues in question were removed. RESULTS. In all cases where papilla were removed reactions still con- tinued. As soon as such animals as had also the nettle welt removed were tested, the experiments were seen to be confirma- tory of those on the papilla. There was always a definite response to light stimulus, but in some casesit was slow. And it is still noticeable that the oral side of the medusa was more sensitive to light than the aboral. It ought to be stated here that the velum was always removed with the welt of tissue and it was not neces- sary to wait for its regeneration in order to get responses to light. To determine if any of the other marginal organs or even any of the organs on the oral side of the animal had anything to do with the light perception, exclusively, or whether, as I began to think, it was the function of more than one organ — perhaps the subumbrellar epithelium in general — other organs which it was thought might be stimulated by light were removed. The ovaries, the velum, and the tentacles were in turn cut away. It is diffi- cult to remove the velum entirely, as the closest cutting that can be done with the finest scissors leaves a narrow bit of the attach- ment of the velum on the margin. However, since the velum may be entirely removed while removing the welt and papille, a differential result will give approximately the value of the velum. Although these operations are more or less severe for the size of the organism, yet I have never observed so severe a shock from any of them as recorded by Yerkes, viz., that they have not again recovered. Indeed, I have almost always removed all the tentacles together and do not remember to have lost any individuals from this cause. A preliminary experiment on a normal one, one without ovaries, one without tentacles, one without the welt and papillz, resulted as follows; For the normal the average reaction time was 74 seconds ; for the one without ovaries 103 seconds ; for the one without tentacles 202 seconds; for the one without nettle welt and papilla 124 seconds. In the light of later observations these results, except the last, seemed too long, as those without ten- LIGHT RECEPTIVE FUNCTION OF GONIONEMUS. Hf tacles or gonads displayed a remarkable activity, probably due to the removal of these organs, In one of the first experiments a normal medusa was compared with two of the same lot in which the marginal nettle welt and papillae were removed. The averages of ten trials were as fol- lows: Average reaction time of the normal medusa 82 seconds ; a 12 mm. olive-colored operated medusa 9} seconds; a 6 mm. yellow orange operated medusa 12 seconds. Next to this in reac- tion time were the medusze with gonads removed. The average of four sets of experiments was 134 seconds. For the velum the average was higher, being 272 seconds ; and for the tentacles still higher being 29seconds. Another case, four hours after removal of the tentacles gave an average of four seconds. DISCUSSION. Now, comparing these, account must be taken of the opera- tions necessary for the experiments. The removal of the tentacles is the simplest of all and least likely to prejudice the action of other parts, therefore these results are the least doubtful. The careful removal of the gonads can have very little detrimental effect, as no principal part of the nervous system is involved. The removal of the marginal papilla may include some of the underlying tissues, and more of this is included when the nettle welt is removed in addition. The same is true when the velum is completely removed. In the experiment for the value of the velum the attaching margin had to be left in order not to injure _ the nerve ring too seriously. In view of these facts negative results after any similar operation, such as Yerkes obtained for the removed papillz, must be looked upon as needing confirma- tion in other ways, rather than being considered a proof of the function ascribed to the organs under consideration. In the specimens without the attachment of the velum, and those with the nettle welt and papilla removed we must bear in mind that parts of the nervous system are more or less affected. CONCLUSIONS. From the earlier experiments of removing the papillz alone it is evident that they are not exclusively the organs of light stim- 8 L. MURBACH. ulation. From the later experiments we may conclude that the welt of pigmented tissue running around the margin of the bell from papilla to papilla is not important, if at all sensitive to light. And furthermore that the ovaries, tentacle and velum have prac- tically no more to do with this function than other organs. In- jury to the nervous system may account for the slow reaction when the margins were removed for marginal welt of tissue and the velum. If now none of these organs that have been tested are solely affected by light, indeed only seem to slow the reaction in proportion to the injury, it seems to indicate that the epithelial tissue on the subumbrellar surface, in general, is the responsive organ. ae PvE AwyY NOTE ON THE DISTRIBUTION OF Peek BEETLES (CICINDELA) AND ITs ime ON LO. PLANT SUCCESSION: V. E. SHELFORD. The adult beetles are graceful, predatory, swift-flying insects whose definite distribution and great variability have long been matters of comment. The larve have been found to be more circumscribed and definite in their distribution than the adults. Our attention has accordingly been turned to the behavior of the larve and of the adults at the time of egg laying for a possible explanation. The egg laying habits are simple. The last four segments of the abdomen are used as an ovipositor. Two pairs of the appendages of these segments serve as digging organs with which small vertical, well-like holes from 7 to 10 mm. deep are made in the soil. A single egg is deposited in a hole and the hole is left uncovered. The tenth segment of the abdomen and one pair of appendages of the ninth segment are covered with hairs which are probably associated with organs sensitive to the varying de- grees of soil moisture and the size of soil particles. The females try the soil before depositing eggs. They make many holes, but lay in only a part of them, and frequently discard them before the usual depth is attained. The larve almost always remain in the spot where the eggs were laid. Upon hatching each larva constructs a burrow in the place of the ovipositor-hole and reconstructs and enlarges the burrow after each moult. If a larva migrates, it almost always selects the same kind of place for digging a new hole as that in which the eggs were laid. So much for the general aspects of the life habits. Let us now turn to some examples of distribution and behavior during the egg-laying time. The larve of Czcindela Ge limbalis are found on steep a IO V. E. SHELFORD. clay banks.’ The range of the adults is far wider. To determine the cause of this distribution, adults were placed in cages contain- ing soil of several kinds. Each kind was so arranged into steep and level parts that about one square foot of each type was ex- posed. The adults placed in the cages were taken when the species was copulating freely. The following table shows the number of larvee which appeared in the case of three lots of C. purpurea imbals. ; Clay and Sand and Soil. Humus. Tee eater Clay. lean Sand. Humist SS Steep, 2 Tevel. |S 290] see hae Sel) ort fete ly )s| Sah eee I. No. of Larvee. fo) fo) fo) fo) 9 @ || @ fo) fo) fo) Il. No. of Larve. Co) fo) fo) fo) 12 I fo) fo) fo) fo) Ill. No. of Larve. fe) fo) I fo) 24 | 10 Co) fe) fo) fo) Other pairs taken in coitus were placed in cages containing sand only and level clay only. No larvz appeared in either case. Females placed in cages containing rough steep clay, deposited eggs. Similar experiments have been carried out on several other species and it becomes apparent, therefrom, that the local distribution is determined by the egg-laying instincts. Since the animals cannot and do not continuously remain far from the breeding place, the breeding place becomes the true index of their habitat. Their local distribution being determined by egg- laying instincts, in other words by the life needs, and housekeep- ing habits of the animals, it may be called ‘ ecological distribu- tion’’ from Haeckel’s’? definition of that term, and the etymology of the word. Habitat selection in correlation with geological factors such as erosion and deposition, and with the succession of plant forma- tions and societies forms one of the great factors of dispersal, iso- lation, etc. 1The nomenclature used in this paper is to be found in Horn’s ‘‘ Systematischer Index der Cicindeliden,’’? Deut. Ent. Zeit., Feb., 1905, Supplement. C. scutellaris Say, however, stands in that publication as obscura Say aber. Lecontez Hald., the corresponding change having been made by the same author in a later publication. 2 Ecology is the science of the domestic side of organic life, of the life needs of or- ganisms and their relation to other organisms with which they live. ‘‘ Wonders of Life,’’ 1905. DISTRIBUTION OF TIGER BEETLES. II The relation of the distribution of Czczzdela to the succession of plant societies has been especially studied in the vicinity of Chicago. The area which affords the basis of this study is to be found at the south end of Lake Michigan. Conditions here since glacial times have led to the deposition of large areas of sand, which in the eastern portion of the field of deposit, is stretched over an area of several miles wide. At the point where most of the studies have been made there is a series of ridges which were originally thrown up under water and later added to by aerial deposition. These ridges are separated by long depressions, most of which contain water. The structures are.accordingly arranged in a horizontal series, the oldest being, of course, furthest from the lake, and differing from the younger only in age, and in being a little less exposed. A definite succession of plant societies has been worked out by the plant ecologists (Cowles and Clements) and this succession is due largely to the conditions necessary for the germination of the seeds and growth of the seedlings of the different plants. In forest development, before the climax stage is reached, the seeds of the trees comprising a given stage do not germinate and their seedlings do not develop in the shade of the forest then present. Each stage accordingly prepares for another and more mesophytic type. The trees of the climax stage of eastern North America, the beech and the maple, produce seeds that will germinate in that forest's own shade. Accordingly the beech and maple will last indefinitely. Not all of the conditions herein described occur in the hort- zontal series at any one point, but all are to be found within sandy areas near Chicago. Let us start with the strip next to the water’s edge, the very youngest deposit. It is frequented by the adults of C. cupras- cens and hirticollis, The larve of the latter are sometimes found here, but more frequently a little further back on the low, wet places on the beach. Other ridges are seen to be formed beneath the water and this margin is accordingly potentially the first. depression. On the lakeward side of the first ridge, among the young cot- tonwoods, we find the larvz of /epida, the white tiger beetle. On the leeward side where bunch-grass has come in and the cotton- 12 V. E. SHELFORD. woods are old with occasional seedlings of gray pine intermixed, we find the larve of C. /epida displaced by those of C. formosa var. generosa, which reach their dominance among the young pines. Coming in on the ridges with the pine are the larve of C. scztel- Jarvis. In our horizontal series this species is to be found further from the lake than any others yet mentioned. As new ridges are thrown up outside of a given one and as it becomes older, the differences between the lakeward and the landward exposure quickly disappear. Let us turn our attention fora moment to depressions. We have noted that C. /irtzcollis occupies a station on the white sand of the beach. In addition to occurring eccasionally in the wet situations just mentioned, the larvez are found in any of the fresh natural depressions that are deep enough to be continually moist at, or near the bottom. Such depressions sometimes occur on our lake shore, behind a first line of small dunes. Asa depression becomes older and the sand becomes somewhat darkened by the decay of the reed, Juncus balticus, the larve of C. hirticollis give way to those of C. repanda which occur a little higher up than the former, on the sloping sides of the depressions. As the /uncus becomes thicker and a few other plants come in, the larve of C. repanda become a little less numerous. We have been able to follow this process in an artificial depression. Finally the vege- tation becomes so dense as to drive out the larve of C. repanda entirely. They are succeeded by the larve of C. tranquebarica which occur still higher up the side of the depression. This stage is coincident with the development of young gray pines on the ridges. Shrubs of various sorts appear on the depression margins at this stage and gradually increase in numbers. The first are the willows and the shrubby cinquefoils. These are succeeded by the button bush and swamp white oaks which make the depres- sion margin too shaded for the larve of C. tranguebarica and the tiger beetle succession of the depression margins is at an end. This stage of the depressions corresponds to the establishment of the black oak, which succeeds the pine, on the ridges. Returning to C. scutellaris at this stage, we find it still, in the DISTRIBUTION OF TIGER BEETLES. 13 open places of the black oak ridges. These oaks are destined not to remain and are crowded out by the coming in of the white oaks. For an immense period after this, the habitats of C. scutcllaris become more and more narrowed. Long before the next tree, the red oak, makes its appearanice, C. scwtellaris has been crowded out. Many centuries must pass between the coming in of the white oak and the establishment of the red during much of which time the Czczndelas are entirely absent. With the establish- ment of the red oak, conditions are ready for the next tree, the shag-bark hickory, and with it comes C. sexguttata. This species appears to reach its dominance in the early stages of the white oak-red oak-hickory forest, and to be crowded into its margins with the development of further mesophytism. It continues in the roads, clearings, fired places, and paths of cyclones in this forest fora long period. Individuals are sometimes to be found in the dense parts about a fallen tree. This type of forest is, how- ever, destined to disappear and its disappearance is heralded by the coming in of the seedlings of the beech and the maple. With their appearance C. sexguttata becomes rare in the forest proper. This species does not deposit eggs in pure humus, but makes use of little irregularities in clay or sand, which, contains a little humus and which is shaded slightly, such conditions as are afforded by falling trees and the erosion of hill sides by small brooks. It prefers a few loose leaves and will lay eggs under them in preference to other places when they are present. It does, not, however, appear to like very shady conditions. Several days spent in the beech and maple forest has failed to reveal the presence of one of these insects although they were present in open and partially cleared places a short distance away where the forest has not become so mesophytic. ; The beech and maple forest is very shady and has a floor of decaying leaves about one inch deep and several inches of very mouldy humus below these, so that there is no place in the forest proper where C. sexguttata can deposit eggs. Itis driven out by the development of these conditions. The white oak-red oak-hickory forest is now distributed over much of the eastern half of northern North America, but the climate in which the beech and maple will develop extends west- I4 Vv. E. SHELFORD. ward only to the Mississippi and Illinois rivers. Dr. Cowles, whose work is still unpublished, has studied the forest of the eastern United States and has come to the conclusion that with the base leveling of the eastern plateau the beech and maple forest would, man eliminated, succeed the less mesophytic types and come to completely cover the territory extending to the western limit of its climatic range. This forest would then come to occupy the entire territory east of the Mississippi and Illinois rivers. This means the driving out of C. sexguttata which is now abundant in the forest of this region. Not at once, to be sure, but irregularly and gradually, first giving irregular and finally discontinuous distribution with a constant narrowing of the range of the isolated habitats. For immense ages habitats would no doubt continue to exist, but since the differences between the different elevations and brook and river margins on the one hand and the climax forest on the other become less and less as the development of the climax stage proceeds, the chances for the maintenance of habitats of C. sexguttata indefinitely, seem small. The general effect of the development of these conditions on the distribution, would be as follows: C. sexguttata would be left only in that portion of its present range, west of the climatic conditions suitable for the development of the beech and maple forest, with possible remnants in the eastern plateau which might by a process of isolation, be caused to take on new habitats and new characters. The general principles here set forth apply to the Czcende/as associated with the development of rivers and the erosion of up- lands. Observations now under way go to show that they apply to the fauna in general. Strikingly different faunas are to be found in the different forest stages herein mentioned. Plant suc- cession is then a factor which we cannot afford to neglect in con- sidering distribution and evolution. LITERATURE CITED. Cowles, H. C. ‘ ?01_ The Plant Societies of Chicago and Vicinity, Geographical Society of Chicago, Bull. No. 2. 1go!. Clements, F E. ’05 Research Methods in Ecology, Lincoln, Neb. 1905. ARTIFICIAL PARTHENOGENESIS IN THE SILKWORM. VERNON L. KELLOGG. I may be pardoned, because of the brevity of this paper, for recalling attention to a subject that seems (but is not) pretty well wor. Really only three men (Tichomiroff, Verson and Quajat) have contributed the data of observation and experiment which have furnished the literature of parthenogenesis with such a host of fleeting references that it must seem to the casual reader as if silkworm parthenogenesis had been investigated only less than that of the sea-urchins. Asa matter of fact it has been investi- gated (the work described in the present notes included) but little. In aclutch of unfertilized eggs oviposited by a virgin silkworm moth (Bombyx mort) almost always a small number of eggs be- gins development. This development extends to the formation of the embryonic envelops and sometimes farther, and is clearly indicated to the observer by the change in color of the egg from yellow to cherry or through cherry to gray. Non-developing eggs remain yellow and, after a while, collapse. Eggs which begin to develop either persist in spherical shape, which indicates persisting life, or collapse, which means death. The development of unfertilized eggs rarely proceeds, without artificial stimulus, beyond a very early embryonic stage. In fully 500 clutches or broods of unfertilized eggs (from confined females from isolated cocoons) under observation, not a single egg gave up its larva, although an average of about seven or eight per centum of the eges began to develop. Although this parthenogenetic development always ceases and the embryo dies before reaching hatching stage, much difference in vitality or duration of life of the egg (strictly, embryo) is notice- able. Some of the developing eggs collapse within a few days, some in a few weeks, while a few persist for several months. (The normal egg stage, 7. ¢., time from egg laying to hatching of larve in the silkworm univoltin races, is about nine months.) ) 16 VERNON L. KELLOGG. There is also to be noted a difference among races in the propor- tion of unfertilized eggs which begin to develop. Among a dozen races in our rearing rooms, one (a vigorous white-cocoon race called Bagdad) is strongly inclined to normal parthenogene- sis, from twenty-five to seventy-five per centum, even in a few cases ninety-five per centum, of the eggs in unfertilized lots be- ginning to develop. The more usual proportion, however, z. ¢., that shown by the other races, is, as already noted, less than ten percentum. So much for normal parthenogenesis in the species. In 1885 Tichomiroff discovered that by bathing the unferti- lized eggs with concentrated sulphuric acid, or by rubbing them gently, he could induce a considerably larger per centum than the normal to begin development. He repeated his experiments, confirming and extending his results, in 1902. By histologic examination of the eggs he learned that the artificially stimulated eges which develop do so in a somewhat abnormal manner. Tichomiroff held the stimulus to development to be neither the action of specific ions, osmotic pressure nor catalysis. He be- lieves that the eggs respond by segmentation to any appropriate excitation, ‘‘ whatever the nature of this excitation.” Verson, in 1899, used electricity as a stimulus, and found that the development thus initiated ceased at a point about corres- ponding with that reached by a fertilized egg on the third day after oviposition. Quajat (1905) submitted unfertilized eggs to the action of oxygen, high temperatures, sulphuric acid, hydrochloric acid, carbon dioxide, and electricity. His account of the experiments indicates that he was able to stimulate development by several of these agents, but he gives no data to show the proportion of developing eggs in the various treated lots. No larva issued, but by an examination of the eggs he found that several embryos had practically completed their development and growth. My own experiments include the treatment of something over a hundred lots of unfertilized eggs (a “lot” is all the eggs laid by a single female, averaging from 100 to 350 in number), and of several lots of fertilized eggs (to serve as controls to indicate possible injury to the eggs from the reagents used). The stimuli or agents used were dry air (obtained by drawing air through vessels ARTIFICIAL PARTHENOGENESIS IN THE SILKWORM. 17 of calcium chloride and then of concentrated sulphuric acid), high temperature, sunlight, friction, sulphuric acid, hydrochloric acid, glacial phosphoric acid, glacial acetic acid, absolute alcohol, potassium hydroxide, ammonia, and lime water. The reagents were used in different dilutions and for varying lengths of time. The treatment was applied to eggs not more than twelve hours old; mostly to eggs but a few minutes to a few hours old. Five hundred or more lots of untreated, unfertilized eggs were observed in order to determine the extent of normal parthenoge- neticdevelopment. The eggs of half a dozen silkworm races were used and all the eggs were preserved from time of laying until their death. As it seemed to me that most of the favorable results obtained by Tichomiroff and Quajat were obtained by treatments which had as common effect a dehydration (such as high temperature, fric- tion, sulphuric acid, etc.) I attempted to test this first by using various dehydrating agents, especially a dry chamber in which the eggs could be submitted for from a minute or two to several hours to a nearly perfectly dry atmosphere. Friction, heat, sul- phuric acid, phosphoric pentoxide and glacial phosphoric acid were also used as dehydrating agents. At the same time other treat- ment, not dehydrating, was used on other lots and gave results hardly less favorable than the dehydrating. The results at the end of this first course of treatment seemed to point to the hydrogen ions as the most likely development-inciting factor. Hence va- rious agents agreeing in containing hydrogen ions though differ- ing radically in other particulars were used. The results gave no encouragement to the hydrogen ion theory. In fact I have not been able to come to an opinion concerning the true causa effictens in the matter. My results simply show to me that various stimuli, acid or alkaline, dehydrating or non-dehydrating, posses- sing or not possessing hydrogen ions, are able to increase mate- rially the proportion of eggs that develop in lots of unfertilized eggs. The following paragraphs give baldly a summary of the results obtained. Treatment of Unfertilized Eggs by Dry Air.— Freshly depos- ited eggs placed in dry chamber for from 14 minutes to 2 hours. Ten lots of unfertilized eggs. In all these lots, except one, a 18 VERNON L. KELLOGG. proportion not exceeding the normal reached the gray stage. In several lots the proportion reaching the cherry (earlier) stage was distinctly above the normal. In one lot two thirds of the eggs reached the gray stage (probably’a lot of Bagdad race eggs). One fertilized lot was treated to see if the drying had any inju- rious effect. Submitted to the dry air for thirty minutes this lot developed normally and all but ten or twelve eggs (a normal number) hatched. Treatment with Sunlight. — Two lots of unfertilized eggs put in direct sunlight for one and two hours respectively (temperature 35° C.). Ten eggs in each lot reached gray stage, a normal number. Treatment by Friction. — Several lots rubbed with tooth brush, not very hard. A small increase over normal average of gray eggs, some of these grays persisting alive for nine months, z. ¢., time for hatching, but none hatched. Treatment by Heat. — Lots heated in oven to various temper- atures from 25° to 57° C. The higher temperatures caused death of all eggs, as well as eggs of fertilized lots used as checks. No increase, over normal average, of developing eggs, through use of the non-fatal temperatures. Treatment by Phosphoric :Pentoxide and Glacial Prosphoric Acid. — Nine lots treated for from one half minute to one hour, the acid applied in some cases as powder, in others as liquid solu- tion. The records are of sufficient interest to give in detail. Lot 1: Treatment one hour. Acid put on as powder. Lot of sixty eggs. Treated June 6 (1906). June 11, three gray eggs. June 21, six gray eggs. August 28, twenty-one or more cherry and gray eggs, of which three are alive, others dead. Lot 2: Treatment, one hour. Acid put on as powder. Lot of one hundred eggs. Treated June 6 (1906). June II, seven gray eggs. June 21, seven gray eggs. August 28, thirty cherry and gray eggs (one half are 29), but all are dead. Lot 3: Treatment, one hour. Acid put on as Re os ot of seventy-five eggs. Treated June 6 (1906). ARTIFICIAL PARTHENOGENESIS IN THE SILKWORM. ige) June II, one gray egg. June’21, one gray egg. August 28, twenty-five grayish-pink and five gray eggs, but mostly dead. March 5 (1907). All eggs are dead. Lot 4: Treatment, two minutes. Acid in concentrated solu- tion. Lot of sixty eggs. Treated June 20 (1g06). June 27, three eggs, partly gray. August 28, eight gray; these and most of the yellow eggs still alive. March 5 (1907), all dead. Mots; yolreatment, two minutes. Acid in concentrated solu- tion. Lot of 250 eggs. Treated June 20 (1906). June 27, five or six gray or cherry eggs. August 28, fifteen cherry eggs; two gray, a few of which are alive. March 5 (1907), all dead. Lot 6: Treatment, one minute. Acid in concentrated solu- tion. Lot of 205 eggs. Treated June 25 (1906). June 27, one grayish egg. August 28, three gray, eleven cherry eggs. Almost all alive. March 5 (1907), all dead. Lot 7: Treatment, one minute. Acid in solution. Lot of 140 eggs. Treated June 25, (1906). June 27, two gray eggs. July 1, fourteen gray eggs, mostly alive. Most of the yellow eggs also alive. March 5 (1907), all dead. Lot 8: Treatment, one half minute. Acid in solution. Lot of fifty eggs. Treated June 27 (1906). July 1, two cherry eggs. August 28, five gray eggs; several cherry, mostly alive. About two thirds of the yellow eggs also alive. Lot 9: Treatment, one half minute. Acid in solution. Lot of ninety eggs. Treated June 27 (1906). July 1, eight cherry or grayish eggs. August 28, twenty-seven gray or pink-gray eggs of which 20 VERNON L. KELLOGG. only two or three are collapsed Of the sixty or more yellow eggs, only six or seven are collapsed. ® March 5 (1907), four live gray eggs; all others dead. The treatment with glacial phosphoric acid seems to have the curious effect of prolonging the life of all the eggs whether they begin actual development or not, and of s/ow/y initiating develop- ment in aconsiderable fraction of them, a proportion distinctly above the average number that would begin development with- out artificial stimulus. Treatment by Sulphuric Acid. — Sixteen lots of unfertilized eggs and two of fertilized (as controls to indicate possible injury by the reagent) were treated with concentrated sulphuric acid for periods varying from one fourth of a minute to two minutes, and then washed with water. This acid is, of course, a strong dehy- drator. In several cases only part of a lot would be treated, the other part left untreated as a check lot. The fertilized eggs developed normally and hatched, showing that the concentrated acid applied for two minutes does not injure the eggs. In all the treated unfertilized lots the proportion, above the normal average, of developing eggs was materially increased. This is also true of the treated parts of lots as compared with the untreated. For example, in lot 2, a large lot of four hundred and fifty eggs, one hundred and fifty were treated and three hundred left untreated. In seven days ninety of the treated eggs were gray, while only five of the untreated eggs were gray. In lot 3, one hundred and forty eggs, one hundred were treated and forty left untreated. In ten days more than half the treated eggs were gray and alive, while none was gray in the untreated part. On the average from thirty to fifty per centum of the eggs in treated lots or fractions of lots began to develop, while in untreated parts of lots the per centum of developing eggs was less than ten. Treatment by Hydrochloric Acid. — Twenty-four lots of unferti- lized eggs treated with concentrated hydrochloric acid or with ten per centum hydrochloric acid, for periods of from one fourth minute to two minutes, then washed with water. The acid has but little dehydrating effect. On the whole the results show the distinctly stimulating effect of the acid, but some lots behaved aberrantly and the proportion of developing eggs did not go ARTIFICIAL PARTHENOGENESIS IN THE SILKWORM. Z1 beyond thirty per centum, and was usually not more than twenty or twenty-five per centum. The eggs treated with concentrated acids for the shorter periods, z. ¢., one fourth and one half minute, were in better condition than those treated for one or two minutes. The eggs treated with ten per centum hydrochloric acid showed no special stimulation. Treatment with Absolute Alcohol. — Killed the eggs. Treatment with Potassium Hydroatde— Eleven lots of unferti- lized eggs treated with strong solution of potassium hydroxide for periods ranging from one fourth minute to two minutes. All the eggs treated were loosened from their resting place and soon collapsed. Before dying the eggs showed a reddish color like the normal cherry of developing eggs, but from the great prevalence of this color in all the treated lots and parts of lots I am inclined to believe this color due to some special effect of the reagent on the egg shell rather than the indication of development. In the lots treated with strong solution for two minutes death and collaps- ing soon occurred, and in the lots and parts of lots treated for one fourth minute with half strength solution, collapsing occurred before it did in untreated lots and parts of lots. Treatment with Lime Water.— Five lots of unfertilized eggs were treated with saturated lime water for periods varying from three minutes to one hour. No increase in proportion of de- veloping eggs. The eggs of a fertilized lot treated with lime water for three minutes; all (except the small normal per centum) developed and hatched. Treatment with Glacial Acetic Acid— Seven unfertilized lots treated with glacial acetic acid, strong and half strong, for periods of one minute. Behavior of the lots uneven. In three of the lots no stimulating effect was noticeable. In two about ten per centum of the eggs developed. In one about thirty per centum began development, while in lot 7, a lot of three hundred and twenty-five eggs, half of which were treated and half left untreated more than fifty per centum of the treated eggs began develop- ment, while in the untreated lot very few, not more than two per centum. In a fertilized lot treated with the acid for one minute, all the eggs, except four or five, developed and hatched. Treatment with Ammonium Hydroxide.— Six unfertilized lots 22 VERNON L. KELLOGG. treated with strong ammonium hydrate for periods of one half minute or one minute. In two lots there was a beginning devel- opment of one third of the eggs ; in the other four lots no increase over the normal average. In one of the two lots showing stimu- lation some of the eggs were left untreated and the increase in "proportion (reaching thirty-three per centum) of the developing eggs occurred only in the treated portion of the lot.. In the un- treated portion only four per centum of the eggs began develop- ment. A fertilized lot treated with the reagent developed and hatched normally. STANFORD UNIVERSITY, CALIF. THE EARLY DEVELOPMENT OF THE LATERAL MINE Svs lM, OF AMIAD CAINA.” CORA J. BECKWITH. The earliest stage of the lateral line system of Asma described by Allis (88) is in an embryo a day after hatching. In this stage the supra-orbital, infra-orbital, opercular-mandibular and the post-auditory lines are well formed but no individual sense organs are differentiated. Wilson (’91) has described for Sevranus a com- mon anlage for the lateral line, auditory organ and branchial sense organ. An elongated furrow formed from the inner layer of the ectoderm lies on either side of the neural tube. This furrow becomes divided by transverse constrictions into anterior, middle and posterior parts. The anterior and middle parts are trans- formed into vesicles by the closing together of the lips of the fur- row and become respectively the branchial sense organ and the auditory organ. The posterior part remains as a groove and is converted into the lateral line. Wilson and Mattock (’97) describe for the salmon a thicken- ing in the form of a solid cord which behaves exactly as the fur- row in Serranus. It is constricted into three parts, the two an- terior of which,form vesicles, while the posterior remains as a cord which is the anlage of the lateral line. Mitrophanow (’93) has described ‘a. similar condition in selachians. In this case there is a shallow furrow which merges gradually into the sur- rounding tissue, thus giving a less distinctly defined groove than in Serranus. The present work was begun with the purpose of tracing the lateral line system in Azzza from its first appearance up to the point where Allis took it up (¢. ¢., in embryos a day after hatching). It was thought possible that the anlage was formed and differentiated in Amza as described for teleosts by Wilson and Mitrophanow. The work was done at the Zoological Laboratory of the Univer- sity of Michigan in 1900, under the direction of Professor Jacob Reighard, to whom I wish to express my sincere thanks. 1 Contributions from the Zodlogical Laboratory of the University of Michigan, No. II4. 23 24 CORA J. BECKWITH. In embryos in which the optic vesicles and hind brain are just being differentiated from the neural tube there is seen in surface view a thickening (Fig. 1, ¢#.) on the neural tube just back of the hind brain. This thickening appears to be a lateral extension of either side of the neural tube and is very similar in appearance to the optic vesicles. It was thought that this might be the anlage for the lateral line. This view was made more probable by the fact that in a later stage (Fig. 3, ¢#.), when the first gill slit (spiracular) is formed, this thickening becomes divided on either side by a transverse fissure into two lobes, the cranial one of which lies opposite the hyoid arch. This bilobed condition sug- gests the division spoken of by Wilson in teleosts. Sections (Fig. 2, z.c.) through the region of this thickening in the first stage described (z. ¢., Fig. 1) show the neural tube still connected with. the ectoderm while a wedge-shaped mass of closely packed cells, or neural crest, lies on either side of the tube between the ectoderm and the dorsal surface of the tube. Later- ally these wedge-shaped masses project out beyond the neural tube. The lines of separation between this very much thickened neural crest and the neural tube on one hand and the ectoderm on the other are very indistinct. This neural crest corresponds exactly in position and extent to the external thickening described for the stage shown in Fig. 1. ; In sections through the anterior half there is on either side of the crest the deep auditory invagination of the inner layer of ecto- derm. The invagination is directed ventrally and inward toward the neural tube. The medial side lies against the broad end of the wedge-shaped neural crest. The wall of the auditory invagi- nation consists of a single layer of columnar cells. The two walls are separated from each other by a plane extending from the outer layer of ectoderm to the inner surface of the invagina- tion (Fig. 2, a.z.). Sections toward the anterior end and middle of the crest show that the auditory invagination ends abruptly in both cranially and caudally. In a little older embryo (Fig. 3) sections show that the neural tube has lost its connection with the ectoderm. The neural crest has become divided into an anterior and posterior lobe as de- scribed in surface view and is not so clearly defined as in the pre- a LATERAL LINE SYSTEM OF AMIA CALVA. 25 vious stage, since the cells are more loosely packed and the cells of the ventral and lateral surfaces merge gradually into the sur- rounding mesoblast (Fig. 4, z.c.). On account of this pushing apart of the cells the crest extends farther ventrally along the sides of the neural tube. Laterally it extends out on either side to a distance equal to the width of the neural tube. The auditory invagination lies between the two lobes of the crest. A cavity has appeared in the auditory invagination in the form of a slit separating the two walls along the plane spoken of in the previous stage. A constriction separating the invagination from the ectoderm is in process of formation. In a stage in which two gill slits (spiracular and first post- hyoidean) are formed (Fig. 5) the two lobes of the neural crest seen in surface view of the previous stage are no longer apparent in surface view. Sections show that the lobes of the neural crest have extended further laterally and at the same time thinned out dorso-ventrally so that they are transformed into two narrow bands of mesoblast extending outward and forward from the neural tube, one into the hyoid arch and the other into the first branchial arch. The auditory invagination lies at the median end of the post-hyoidean slit between the ends of the hyoid and branchial arches and extending into the mesoblast of the pos- terior edge of the hyoid arch. The constriction observed in the last stage between the ectoderm and the auditory invagination has proceeded until the auditory invagination is now a closed ves- icle nearly separated from the ectoderm (Figs. 5, a.p.; 6, @.27.). This auditory invagination was traced in sections of later stages until in a stage like Fig. 7 it has become a large vesicle, circular in section, lying some distance below the ectoderm. Its median half is partly covered by the hind brain so that only the lateral portion is visible in surface view. Its wall is composed of a single layer of cells which are more columnar on the median side than on the lateral. The thickening on the neural tube proves then to be a mass of mes-ectoblast in the form of a neural crest in which the audi- tory invagination lies and so cannot be the anlage of the lateral line system. The neural crest of this region divides into two lobes which go in large part to form the mesodermal portion of the 26 CORA J. BECKWITH. hyoidean and the first branchial arches. It also probably gives rise to the seventh and ninth cranial nerves. In the region of the cranio-lateral portion of the neural crest there is on each side the auditory invagination of the ectoderm. This becomes con- stricted from the ectoderm and forms aclosed vesicle or auditory vesicle. As the auditory vesicle undergoes no further division it is evident that the lateral line anlage does notarise in connection with the auditory organ as described in teleosts. The beginning of the lateral line system appears in a later stage. The first indication of a lateral line is found in sections of a stage long after the establishment of the auditory vesicle (Fig. 7) and about a day and a half older than the stage represented by Fig. 5. A thickening of the inner layer of ectoblast caused by the cells becoming columnar runs along each side of the embryo in the angle which the embryo makes with the yolk (Fig. 8, f.@./.2.). It extends from the second gill slit to the region of the first somite. It is not visible in surface view. This proves to be the anlage of the post-auditory division of the lateral line system. In embryos about a day and a half older than that last de- scribed the lateral line is first seen in surface view (Fig. 9, p.a././.). The post-auditory anlage has elongated so as to extend to the middle of the second post-auditory somite. It is a very slender cord formed from the ectodermal thickening described in the pre- vious stage. It is of greater depth than in the earlier stages, so that its outer surface is slightly raised above the surface of the body. The anlage of the head lines also appear in this stage (Fig. 9, 5.0.2.2, 2.0.2.2., 0.ml2.). The supra-orbital and the oper- cular-mandibular lines form a V, the arms of which extend craniad from a point just in front of the auditory organ. The supra-orbital extends to the caudo-dorsal portion of the eye and consists of two elongated bead-like thickenings end to end. The opercular-mandibular extends down the cranial edge of the gill cover and is a very slender line. Passing from the caudo- ventral portion of the eye and into the angle of the V is the infra- orbital line which is a short broad thickening. The lines in the head region are formed by thickenings of the ectoblast which pro- ject inward very slightly, but also raise the surface in the form of ridges. The thickenings are caused here also by the ecto- derm cells becoming columnar (Fig. 10, 2.0. and s.0./..). LATERAL LINE SYSTEM OF AMIA CALVA. 27 In an embryo that has just hatched (Fig. 11) the supra-orbital line extends cranially to a point just dorsal to the eye and isa continuous cord instead of being beaded. The infra-orbital line extends below the eye to the ventral surface of the nasal pit. The opercular-mandibular has changed but little from the condi- tion in the previous stage. The post-auditory line has grown craniad so that the anterior end extends to about the middle of the auditory pit. Caudally it extends to a point half way between the operculum and the pectoral fin. This line consists in this stage of three elongated bead-like thickenings placed end to end. The connection of the post-auditory line with the infraorbital line shows first in an embryo a day after hatching (Fig. 12). This is the stage with which Allis begins his description. The post-auditory line extends caudally to the pectoral fin and shows a fine beaded appearance. The head lines are very much broader than before and the supra-orbital extends farther over the eye. The infra-orbital sends a small twig up over the nasal pit. The line connecting the infra-orbital with the post- auditory line is a very fine cord. A very small twig is given off from this line just below the auditory organ and projects half way up the caudal border of this organ. There is no differ- entiation of the lines into definite sense organs in this stage. SUMMARY. 1. In very young embryos there is present a thickening on the neural tube which consists of a mass of mes-ectoblast of neural crest origin. In this the auditory invagination lies embedded. This portion of the neural crest later becomes divided into two lobes between which the auditory organ lies. In a still later stage the two lobes have become thinned out into sheets of mesoblast which extend one into the hyoid gill arch and the other into the first branchial gill arch. The seventh and ninth cranial nerves are probably also formed from the neural crest. The in- timate relation of the auditory invagination to these lobes of the neural crest produces a striking resemblance to the condition found in the sensory anlage in teleosts. Here a single elongated thickening is described as dividing by two transverse fissures to form the anlage of the branchial sense organ, the auditory organ and the lateral line system respectively. 28 CORA J. BECKWITH. 2. The auditory and lateral line organs of Azza do not have a common anlage. The auditory organ arises before the lateral line and independently of it as an invagination of the nervous layer of the ectoderm and forms later a closed vesicle. 3. The lateral line appears first in an embryo in which the auditory organ is a closed vesicle. The four primary lines (supra-orbital, infra-orbital, opercular-mandibular, and post-audi- tory) all arise independently of one another and of the auditory organ as thickenings of the nervous layer of the ectoderm and unite later to form a continuous system. LITERATURE. Allis, Edward Phelps, Junior. 88 The Anatomy and Development of the Lateral Line System in Amia cals Journ. Morph., II., 463-458, Pls. XXX.-XLII. Wilson, H. V. ’9t The Embryology of the Sea Bass (Serranus atrarius). Bulletin U. S. Fish. Com. Washington., IX. (1889), 209-277, Pls. LXXXVIII.—CVII. Wilson, H. V., and Mattock, J. E. ’97 The Lateral Sensory Anlage in the Salmon. Anatomischer Anzeiger, XIII., 658-660, two figures. Mitrophanow, Paul. ’93 Etude Embryogénique sur les Selachiens. Archives de Zool. Exp. (3), I 160-220, Pl. IX.—XIV. 30 CORA J. BECKWITH. EXPLANATION OF PLATE I. Fic. 1. Embryo of Amza in which there is a thickening (¢.) of the mesectoderm or neural crest on the neural tube just back of the hind brain. 20. Fic. 2. Section through thickening along the line a—é, Fig. 1 showing neural crest and auditory invagination. a@.z., auditory invagination; ec., ectoderm; e7., entoderm ; .c., neural crest; .¢., neural tube. > 167. Fic. 3. Embryo of Aza showing the division of the thickening (¢%.) into two lobes. 20. Fic. 4. Section along the line a—d, Fig. 3, showing the gradual merging of the neural crest into the surrounding mesoblast. ec., ectoderm; ez., entoderm ; 7.c., neural crest; 2.¢., neural tube. > 167. Fic. 5. Embryo of Amza showing the extension of the neural crest into the spiracular and first branchial arches. ad.org., adhesive organ; a@.f., auditory pit ; rst b.a., first branchial arch; %.@., hyoid arch; m.a., mandibular arch. 20. Fic, 6. Section through line a—é, Fig. 5, showing the constriction of the auditory invagination from ectoderm. a.zz., auditory invagination; ec., ectoderm; ev., entoderm ; z.¢., neural tube; z¢c., notochord. < 167. PLATE | 32 CORA J. BECKWITH. EXPLANATION OF PLATE II. Fic. 7. Embryo of Amza in which the lateral line anlage is first visible in section but not in surface view. a@.f., auditory pit. < 20. Fic. 8. Section along line 2-4, Fig. 7, showing the first appearance of the anlage of the lateral line. ec., ectoderm; z., heart; z./., neural tube; z¢c., notochord ; p.a.t./,, post-auditory lateral line. > 167. Fic. 9. Embryo of 4mia showing the auditory pit. The lateral line anlage is visible as ridges, in both head and body regions. a.Z., auditory pit; z.0././. infra- orbital lateral line; 0.7././., opercular mandibular lateral line; f.a././., post-audi- tory lateral line; s.o./.7., supra-orbital lateral line. >< 20. Fic. 10. Section along line a—é, Fig. 9, showing the thickening of the inner layer of ectoderm to form head lines. ec., ectoderm; 27.0././., infra-orbital lateral line ; n.t., neural tube; of.v., optic vesicle; s.o././., supra-orbital lateral line. >< 167. oO To's Aloe y Aaa Fe ~ mh Nip Sis {| 3LV1d AIX "10A ‘NILATING IvoOIDO 1018 34 CORA J. BECKWITH. EXPLANATION OF PLATE III. Fic. 11. Embryo of Amia, newly hatched, showing further development of the four primary lateral lines. @.f., auditory pit; z.0././., infra-orbital lateral line ; o.m.t.d., opercular-mandibular lateral line ; .a././., post-auditory lateral line ; 5.0.7.2, supra-orbital lateral line. X 20. Fic. 12, Embryo of Amza, one day after hatching, showing the four primary lateral lines before definite sense organs are differentiated. >< 20, a.p., auditory pit; z.0./.2., infra-orbital lateral line; 0.7./.7., opercular mandibular lateral line ; p.a.l.l,, post-auditory lateral line; s.0././., supra-orbital lateral line. Il SLvid a) fie bee DING HABITS OF THE” KAINBOW DARTER (ETHEOSTOMA CCERULEUM SOIR), va SVR UID NG SUNS Sica Uae SELECTION. CORA D. REEVES. Contributions from the Zoélogical Laboratory of the University of Michigan. No. 113. CONTENTS. PAGE [Mitre duction cee vomteeacesedes cad nen cases ace tendebeessocnssetes Becca! (sieuinaatee 35 II. Usual Appearance, Habitat, and Behavior of the Fish........ ............008 36 III. The Breeding Colors, Habitat, and Behavior................0206 seeceeee ceceee 38 TON FAVES PA WANN Ohi risrocioee se ielsnis Sslclaslalewisiet sisie'sie eitia avis vise ne /oaearey oldeisiesle@aescioaenioactes Seslseweas 41 1. Behavior of a large male toward a female................0:ceseeeeesencees 42 Zuber Spawminya tite. tenses omasessteccmsescereomsicoe ace cranee avewscieltcsiees 43 Bs hexS pawmingevACtis ccktiscesessa. seussscesmssssceae cereus cncnccwistnelssvseeeoer 44 4. Behavior of two rival males toward each other...............-+-eceeeeees 45 5. Behavior of supernumerary males toward the spawning pair........... 46 OlM SCNEKE COMMU OMEMA Mea nen ctnmecn: scaun ote atere coal maninc seme meee cess macaa dee sacs ac 47 V. Observations bearing on Sexual SelectionS..............cececeseeeecececaececeeee 47 Walter ID ISCUSsionoleNesulltsy ase «ccnsancr