f 3 ge tesa aod te Peter cay “e y file : aed vn ; a : ; "e wa : ‘ a , i ’ ‘ » Fi ae | : yas , - ae othe t = oe ate we ‘ ‘ ‘ art . res - 2 Dy s) a , ; , : : ae Hus thheepyes «i n os pels ’ : ‘ ‘ oe 74 a : a Pete seiiny pes Hop eR Cees bias bal ore po PE DEN Ft ek h oo : Ne = y 7 Pe 4 ‘ f ‘ , a ‘ u ’ ' ji ’ \ ; ‘ A, . l : , , , oe Pox id ‘ sot 4 ; i J : ‘ ia wth. We a he a ue Say ek me rw AUR vitae! Y bene bea EW d Achigee #ACy . ear te aida segue pasta we wend? ioe nh BANS ye ’ he ' tprty 3h Rios hes Wal idathe Reise neste be mst aan aaa st Porat hd Sy, be. set cota: bier > 5 sae AR erg! es wi teen wr es ee Ree ate eh ‘ re walks nal ony Ley aN ehh gare) ne hs anit i San tik a . vast iets So TS ib in rier oh sh Besta area nit Since Is reo itt — 2 ce oH he i .e) . pea) Tet ERS ir bred 4a sn AE SN Pai % ies i aes Veep dincauh fem Thi te We eget Seni ny apy, ay i Mey 4 r¢ (6) SSF PROCEEDINGS OF THE AMERICAN PHILOSOPHICAL SOCIETY HELD AT PHILADELPHIA FOR PROMOTING USEFUL KNOWLEDGE. VOL. XLII. JANUARY TO DECEMBER, »° MX 1904. ss) \\ \ PHILADELPHIA : THE AMERICAN PHILOSOPHICAL SOCIETY. 1904, ee - re PROCEEDINGS AMERICAN PHILOSOPHICAL SOCIETY HELD AT PHILADELPHIA FOR PROMOTING USEFUL KNOWLEDGE Vou. XLII. January, 1904. No. 175. Stated Meeting, January 1, 1904. President Smrru in the Chair. - A communication was received from the Société Nationale des Antiquaires de France, announcing that it would celebrate its centenary on April 11, 1904, and inviting the Society to appoint a delegate to represent it on this occasion. The invitation was accepted, and the President subsequently appointed the Marquis de Nadaillac to represent the Society at the celebration. The Judges of the Annual Election of Officers and Councillors reported that an election had been held on the afternoon of this day, and that the following named persons had been elected to be the Officers for the ensuing year’ President Edgar F. Smith. Vice-Presidents George F. Barker, Samuel P. Langley, William B. Scott. Secretaries. I. Minis Hays, Edwin G. Conklin, Arthur W. Goodspeed, Morris Jastrow, Jr. 4 MINUTES. [Jan. 15, Treasurer. Henry LaBarre Jayne. Curators. Charles L. Doolittle, William P. Wilson, Albert H. Smyth. Councillors to serve for three years. Richard Wood, Henry Carey Baird, Samuel G. Dixon, J. G. Rosengarten. Stated Meeting, January 15, 1904. President SmirxH in the Chair. The decease was announced of Alfred R. C. Selwyn, LL.D., at Vancouver, B. C., on October 18, 1902, xt. 78. The following papers were read: : “On the Phylogeny of the Edentata,” by Prof. William B. Scott. “Herder and Franklin,” by Prof. M. D. Learned. “The Main Facts in Regard to the Cellular Basis of Hered- ity,” by Prof. Thomas H. Montgomery, Jr. 1904.) MONTGOMERY—CELLULAR BASIS OF HEREDITY. 5 THE MAIN FACTS IN REGARD TO THE CELLULAR BASIS OF HEREDITY. (Contributions from the Zoological Laboratory of the University of Texas, No. 56.) BY THOMAS H. MONTGOMERY, JR. (Read January 15, 1904.) ¥: Under heredity we understand the transference to the offspring of qualities of the parent or parents. ‘The interpretation of the phe- nomena involved constitutes one of the broadest problems in the field of Biology, and has for centuries been the theme of eager dis- cussion. Yet only in the past forty years has there come out any positive knowledge upon the subject, except the making known of certain cases of parthenogenesis and of the occasional difference of reciprocal crosses. . There are obviously two methods of determining the facts of heredity. First, by the intercrossing of different varieties or species, and the determination of the relative influences of the parents upon the offspring. The first fundamental work in this line was done by Mendel in 1865 (Versuche iiber Phlanzenhybriden), who determined a large series of facts for the plant genus Pisum, and from the data established a mathematical law for this genus as to the inheritance of parental qualities by the hybrids. This memoir, only some three years ago resurrected from its long obscurity, is to-day occupying the attention it deserves, and has stimulated much work along the same line. De Vries’ magnifi- cent work,. Die Mutationstheorie, demands as well recognition in this respect. But itis clear that such experimental intercrossing, in so far as only the end results of the crosses are considered, can do no more than state the degrees of resemblance of the offspring to the parents, and decide the questions as to the fertility of the hybrids. Important and necessary as it is, it does rot go to the root of the matter, and cannot present any empirical analysis of the underlying factors. For an understanding of these we must turn to the second method, to the examination and interpretation of the intimate structural and growth phenomena of the germ cells themselves, that is, to the 6 MONTGOMERY—CELLULAR BASIS OF HEREDITY. (Jan. 15, cellular basis. All explanations must remain purely hypothetical until this is done. And here I would call attention, as briefly and concisely as possible, to certain positive results that have been won in the study of the germ cells, and disregard the many fascinating but purely hypothetical views as to the process of heredity. II. The statement of the problem must bea very broad one. The fertilized egg gradually cleaves into many cells. These progressively arrange themselves into tissues, and these form organs. By con- tinuing cell division, by change of position and infolding of cells, and particularly by a differentiation of the cells as the development proceeds, the adult organism eventuates. Then from the body of this adult comes an egg, and it repeats the whole involved process. Here are two great fundamental problems: the one, why the offspring resembles the parent? the other, what are the factors of differentiation ? On the answer to these problems depends to great éxtent the explanation of how variations arise and how they are promulgated, that is, the explanation of descent with modification, ‘broadly called evolution. The very subsidiary question of the determination of sex is necessarily also connected with these prob- lems. And all of these questions are inseparable from the one: How far is the adult preformed or prelocalized in the germ cells ? What interests us immediately are the two points: First, has there been empirically determined a particular cellular substance, most intimately connected with the transmission of hereditary growth energies? And second, if such a substance is known, does its behavior during the process of development of the embryo throw any light upon the processes of heredity ? III. To make the following argument clear, we must call to mind the structure of the mature germ cells and the process of cell division. The maternal germ cell, the ovum, appears much like any large, unspecialized cell. We distinguish in it a central rounded body, the nucleus, with its surrounding cytoplasm, In the cytoplasm there is a living substance, the protoplasm proper, and various deu- toplasmic substances, such as yolk, which serve mainly for the nourishment of the cell. The nucleus is more complex. Travers- 1904.) MONTGOMERY—CELLULAR BASIS OF HEREDITY. 7 ing the thinly fluid nuclear sap, which fills it, is a delicate network or,meshwork of linin threads, and supported upon or imbedded in them masses of a substance called chromatin. In the nuclear sap may be suspended also one or many large rounded bodies, the nucleoli, and numerous minute lanthanin granules. The whole is enclosed by a nuclear membrane. The paternal germ cell, the spermatozoon, has a very different appearance, and in volume is exceedingly smaller than the ovum. In the case of the sea-urchin, Wilson (Zhe Cell in Development and Inheritance) has computed it to be about one half-millionth the volume of the egg, and the difference is many times greater than this in the case of the bird. The history of its formation shows it to be a highly specialized cell with regard to its cytoplasm, which is generally modified to form a locomotory flagellum. But its amount of chromatin is the same as that in the egg cell, though contained in a very condensed form (composing the head of the spermatozoon). At the junction of the flagellum and head there is frequently found a mid-body, a metamorphosed centrosome. Thus there is a division of labor between the two germ cells: the ovum is large to provide the necessary cytoplasm and nourishment for the embryo ; the spermatozoon minute and motile in order to reach the ovum. All cell reproduction is by division of the cell, and the mode of division, which differs very notably from a mere constriction into two, may be briefly recalled. The nucleus of the cell increases in volume, and its scattered chromatin masses group themselves evenly along the linin threads, so that eventually the chromatin seems to be arranged in the form of a long, continuous loop. In the cyto- plasm at one side of the nucleus appears a minute body, the centro- some. This divides into two centrosomes, and they wander apart from each other, each through an angle of 90°, to opposite sides ot the nucleus. These centrosomes are the dynamic centres of the cell division and exert an influence upon the surrounding cyto- plasm, as shown by systems (asters) of cytoplasmic rays converging upon them. Within the nucleus, meanwhile, the chromatin loop has become split through its entire length by an exact halving of each of its larger chromatin masses, and has also broken trans- versely into a fixed number of segments, the chromosomes, which now are connected together only by linin threads. Then the nuclear membrane dissolves away and a dicentric figure appears 8 MONTGOMERY—CELLULAR BASIS OF HEREDITY. (Jan. 15, with a centrosome, the centre of an aster, at each pole, the chro- mosomes grouped together in a plane midway between the poles and with the long axis of each chromosome coinciding with this plane. Then begins the separation from each other of the halves of each longitudinally split chromosome and to opposite poles, probably due to the contraction of linin fibres that connect the chromosomes with the centrosomes. Their separated halves come to lie in two groups, one near.each centrosome. Finally, each centrosome loses its influence upon the cytoplasm, the radiations around it disappear, each group of chromosomes forms again a rounded nucleus, the cell body constricts between them to form two cells, and as a result there are two cells each with its own nucleus. The remarkable accomplishment is an exactly equal dis- tribution of the chromatin mass to the daughter cells by a very complex mechanical process. IV. Now is there any particular one of these structures that can be determined as the bearer of hereditary qualities? No one has advocated that it might be a centrosome, and, indeed, there is no reason for considering a centrosome to be any other than a dynamic . centre. Such a substance must then be in either the cytoplasm or the nucleus. The earlier views were that this particular substance was located in the cytoplasm (Lankester, 1877; Whitman, 1878; Flemming, 1882; Van Beneden, 1883). But these were hypothetical assump- tions and employed not so much to show a special hereditary sub- stance, as rather to explain the progressive specialization of the cleavage cells. Hereditary traits cannot, moreover, be transmitted by the cytoplasm of the spermatozoon, for in some cases (Echino- derms) the whole cytoplasmic flagellum of the spermatozoon is left outside the egg, and only the head and midbody of the spermato- zoon penetrate the egg in fertilization. There is also the decisive experiment of Boveri, to which we shall recur, showing that the cytoplasm of the egg cell also does not transmit hereditary traits. Accordingly the hereditary substance must have its seat in the nucleus, and there is now practically positive evidence that such a germ plasm is the chromatin, The main reasons are as follows : (1) The exact distribution of the chromatin in cell division, so that each daughter cell receives just half the amount of chromatin a 1904.] MONTGOMERY—CELLULAR BASIS OF HEREDITY. 9 of the mother cell. The longitudinal splitting of the chromosomes is an autonomous act, whereby each small chromatin mass compos- ing the chromosome (though not the smallest visible granules or microsomes) divides exactly into halves, and the whole complex series of changes leading to the dicentric division figure seem to have been evolved simply to effect the equal distribution of the daughter chromosomes to the daughter cells. Whether the cyto- plasm divides equally or unequally, the chromatin is always divi- ded and distributed equally. This fact alone has seemed sufficient to most workers to mark the chromatin as the hereditary substance. (2) The fact that the chromosomes, the accumulations of chro- matin during cell division, are fixed in number for all the cell gen- erations of a species. And thé strong probability, amounting almost to a fact, that the chromosomes preserve their individual continuity from generation to generation, notwithstanding their great chemical and structural changes during the rest stage of the cell. (3) The fact that the spermatozoon, in most respects the very antithesis of the ovum, on entering the egg in fertilization brings in just the same amount of chromatin as that contained in the egg. Not only is this so, but Van Beneden demonstrated as long ago as 1883 (Recherches sur la maturation de leuf) that the spermato- zoon brings into the egg just as many chromosomes as are con- tained in the latter. Since we know that the two parents have an approximately equal iyfluence upon the offspring, and since the chromatin is a substance contributed in equal amount by the two germ cells, it is logical to conclude that this substance is the seat of the hereditary growth energies. (4) The fact that, despite considerable differences in other respects in their cell divisions, animals and the higher plants show essentially the same behavior of the chromosomes. (5) The experiment, first made by Boveri, 1895 (Ueber die Befruchtungs- und Entwickelungsfihigheit kernloser Seeigel-Eier), of fertilizing with a spermatozoon the cytoplasm of an egg cell deprived of its nucleus. Such a fertilized egg fragment develops, but shows purely parental characters, probably because all mater- nal chromatin had been eliminated. And two recent papers by Boveri (Ueber mehrpolige Mitosen, etc., 1902; Ueber den Einfluss der Samenzelle auf die Larvencharaktere, 1903) have shown, with ; ae 10 MONTGOMERY—CELLULAR BASIS OF HEREDITY. [Jan. 15, their keen critical analysis of the experiments, that the chromatin alone can be considered the bearer of the hereditary traits. From all these results it is concluded that the chromatin is the seat of the hereditary growth energies.* And from another point of view this is rendered probable. The microchemical study of the cell has shown that the chromatin is the most active substance concerned in cellular metabolism : and experimental work, particularly that of Verworn, shows that a cell deprived of its nucleus, and hence of its chromatin, is unable to build up new substances. The chromatin accordingly, as it is transmitted from generation to generation, carries with it certain definite metabolic energies characteristic of the species. And from this view there is good reason to consider the idea of Delage (Za structure du protoplasma et les theories sur I Hérédité, 1895) to be in the main correct, namely, that the offspring is like the parent because it has similar metabolic energies. Vi There is another series of facts known about the behavior of the chromatin, the hereditary substance, in the germ cells, and a few of them will be touched upon. Oscar Hertwig showed, in 1875 (Beitrége zur Kenniniss der Bildung, etc., des tierischen Eies), that the fertilized egg cell contains two nuclei, one belonging to the egg cell itself and one introduced by the spermatozoon. Then Van Beneden (¢. ¢.) demonstrated that the spermatozoon brings in just 1It has been argued by an English writer whose name escapes me, as does the title and date of his paper, that the linin is the hereditary substance. Active chromatin is never disassociated from linin, but there is always a sub- stratum of linin in each chromosome, and in the rest stage the chromatin is always supported upon linin strands, Hence it was argued that the linin is like- wise equally distributed in cell division, This is a good point, but there is a strong objection to it. When the daughter chromosomes separate, in the ana- phase, the linin becomes pulled out between every two corresponding chromo- somes as a connective fibre, and in the reconstruction of the daughter nuclei the greater portion of such a fibre is not taken up again into the nuclei, And this fact cannot be used in favor of the intracellular pangenesis theory of de Vries, whereby pangenes are hypothetically supposed to wander out of the nucleus and so determine the differentiation of the cleavage cells, for the connective fibres appear to behave alike in all cell divisions. Thus of the two constituents of the chromosomes, at each cell division some of the linin becomes displaced into the cytoplasm, but all the chromatin passes into the nucleus, 1904] MONTGOMERY—CELLULAR BASIS OF HEREDITY. 11 as many chromosomes, and that their mass is the same, as those con- tained in the egg. Further, it is proved that in normal fertiliza- tion only one spermatozoon enters the egg, and that when more than one enters the development is abnormal. The proof that both egg chromosomes and sperm chromosomes have an approxi- mately equal réle in determining the growth of the embryo has been shown by Boveri (/. ¢.) by crossing different species of sea- urchins, and by analyzing the results of fertilizing an egg with two or more spermatozoa. Now each act of fertilization would necessarily double the nor- mal number of chromosomes, since the spermatozoon introduces as ~ many as are already present in the egg, were there not some pro- cess to obviate this. There is such a process, and it is known as the ‘¢ reduction in number of the chromosomes.’’ ‘The last two divis- ions of the germ cells, preceding the act of fertilization and pre- paring them for it, are known as the maturation divisions; and it has been known for some fifteen years past that in these divisions each germ cell has only one-half the normal number of chromo- somes. It is also proven that the ripe egg cell, as well as the ripe spermatozoon, has only one-half the number of chromosomes char- acteristic of the species. It is further known (since the work of Henking and of O. Hertwig, in 1890) that the processes involved in producing this result are essentially the same in both germ cells. Accordingly, by this preliminary halving of the number in each germ cell before fertilization, the germ cells on conjugation each contribute only one-half the normal number, with the result that the normal number is restored. But this preliminary reduction in number has a broader meaning than this. Before the first maturation division of the germ cell is accom- plished there takes place a pairing of the chromosomes, so that instead of, ¢.g., four single (univalent) chromosomes there are two double (bivalent) ones (Montgomery, Spermatogenesis of Peripa- tus, 1899). These become so arranged that one of the two matu- ration divisions results in separating chromosomes that are split longitudinally, just as in any other cell division; but the other maturation division removes entire chromosomes from each other by separating the two chromosomes of each pair, and thereby reduces the number of the chromosomes to one-half. That is definitely known for certain species. 12 MONTGOMERY—CELLULAR BASIS OF HEREDITY. {Jan. 15, But how account for the preliminary pairing of the chromo- somes? It is apparent that each spermatozoon may be called paternal, but not male, and each egg cell maternal, but not female, for the following reason: We have seen that each organism formed by fertilization has a fixed number of chromosomes, half of which were derived from the spermatozoon and half from the egg cell. The germ cells that develop within that organism, be they sperma- tozoa or egg cells, accordingly have an equal number of chromo- somes from each parent. Therefore, the spermatozoon contains maternal as well as paternal chromosomes, and the egg cell pater- nal as well as maternal chromosomes, ' And, therefore, each germ cell has in equal measure the hereditary substance of both its parents. Now the process of pairing of the chromosomes, which we found to be an initial step to their reduction in number, has been proved to be a pairing of paternal with maternal chromosomes (Mont- gomery, 4 Study of the Chromosomes of the Germ Cells of the Metazoa, 1901). Ina particular generation of the sperm cell it was demonstrated (and not merely ‘‘ surmised,’’ as stated by another worker) that paternal chromosomes pair with maternal ones, form- ing thus double rods instead of single ones; it is probable, but not yet demonstrated, that likewise in each egg cell, of the correspond- ing generation, paternal chromosomes pair with maternal. Thus in the reduction division, which displaces the two elements of a pair, a maternal chromosome separates from a paternal in each pair, but not so that all the paternal chromosomes pass into one cell and all the maternal into another. These facts which we have learned about the chromatim lead to aconclusion that for its probability approaches a fact. That is, that the different chromosomes in a germ cell have each their par- ticular values. Roux (Ueber die Bedentung der Kernthetlungs- Jfiguren, 1883) was the first to postulate that the chromatin cannot be hereditarily the same throughout the length of a chromosome, for otherwise its equal longitudinal splitting would be without mean- ing. In other words, each particular portion of a chromosome would represent a particular hereditary value. Not only is this probable, but it is also probable that one chromosome has heredi- tary values not found in the others. For we have seen that each germ cell hasa set of maternal and a set of paternal chromosomes, and that ina particular generation those of the one set pair with those 1904.] MONTGOMERY—CELLULAR BASIS OF HEREDITY, 13 of the other (Montgomery, 2 ¢.; Sutton, Zhe Chromosomes in Heredity, 1903). The two that pair are of corresponding vol- ume (as brought out especially by Sutton), and sometimes of cor- responding form (Montgomery, in a paper now in press). Because they are thus similar in volume and form, it is at least possible that they are similar in hereditary value. So Sutton has ably argued that when the two of a pair, a maternal and paternal chromosome of corresponding volume, separate from each other in the reduc- tion division, chromosomes of like hereditary quality become sepa- rated into separate cells, so that no mature germ cell shall contain before fertilization two chromosomes having similar hereditary values. And this is the best reason yet given in explanation of the peculiar reduction division. VL Finally, we may ask how far these facts agree with the germ- plasm theory of Weismann. Some eighteen years ago, Carnoy (La cytodiérése chez les arthro- podes, 1885) showed, and he was the first to do so, that two kinds of cell division occur, namely, a transverse splitting of the chro- mosomes and a longitudinal splitting. ‘That transverse splittings of chromosomes should occur was directly opposite to the prevalent view of the time, to the effect that only longitudinal divisions take place. Carnoy was far ahead of his day, and while this most im- portant memoir of his then and for years afterwards met with only rather scornful criticism, we must now grant him his proper place as the discoverer of the reduction divisions. Weismann, in 1887 (Ueber die Zahl der Richtungskirper und ueber thre Bedeutung fiir die Vererbung), prophesied, clearly with- out knowledge of Carnoy’s work, and in conformity with the ideas of Roux (1883, /. ¢.), that in addition to the longitudinal splitting of the chromosomes, the ‘‘ hereditary equal division,’’ there would be found to occur in certain generations of the germ cells a ‘hereditary unequal division,”’ either by a transverse division of the chromosomes or by a separation of entire chromosomes from each other. A number of the students of the maturation phe- nomena of the germ cells have empirically demonstrated this. Weismann’s reduction division is the one where entire chromo- somes become separated from each other. Equally, confirmation has been brought of another of his cardinal postulates, the con- 14 , MINUTES. [Feb. 5, tinuity of the germ plasm. To be sure it is known that the germ plasm, the chromatin, is not an eternally unchangeable substance, as Weismann at first postulated. But the chromatin persists from generation to generation ; the continuity of the germ plasm is what to-day is being termed the continuity of the chromosomes, and these continue from generation to generation, maintaining their individuality, just as much as a particular cell of one generation may be said to be represented by a particular cell of another. Only some half dozen years ago, in the course of the conflict over the germ-plasm theory of Weismann, no workers upheld the occur- rence of the reduction division except the school at Freiburg and one or two others. There even appeared a paper, presuming to be ‘decisive, entitled ‘‘ The Facts of Chromosome Reduction versus the Postulates of Weismann’”’ (J. E. S. Moore, 1897). Since that time there has been much new research and by the comparative method, perhaps the safest of all methods, and the mass of evidence: is now strongly corroborative of Weismann’s two cardinal postu- lates. So to-day Weismann can point to the actual confirmation of the fundamental portion of his germ-plasm theory. Stated Meeting, February 5, 1904. President Smiru in the Chair. The following papers were read: “The Babylonian and Hebrew Accounts of the Creation, in the Light of Recent Criticism,’’ by Prof. Morris Jastrow, Jr. i “The Miocene Diabase of the Santa Cruz Mountains in San Mateo County, California,’ by H. L. Haehl and Ralph Arnold, communicated by Prof. J. C. Branner, 15 [1904. HAEHL AND ARNOLD-—THE MIOCENE DIABASE. MAP KS ~ wv Ku ~ Showi istribution he: "ee ag Bot 5 MIOCENE i | /) DIABASE i of the : . Lat oe . L—"| ISANTA CRUZ MOUNTAINS in San Mateo Co LAL. ¢"= IB RT OR Miles We ipiobase Basalt Fees oon be ~ & terra Morena {rn 16 HAEHL AND ARNOLD—THE MIOCENE DIABASE, [Feb. 5, THE MIOCENE DIABASE OF THE SANTA CRUZ MOUN- TAINS IN SAN MATEO COUNTY, CALIFORNIA.’ BY H. L. HAEHL AND RALPH ARNOLD. (Read February 5, 1904.) INTRODUCTION.” The presence of the basic eruptives in the Santa Cruz Mountains of San Mateo County, California, and portions of Santa Clara and Santa Cruz Counties, was first noted in 1865 by J. D. Whitney. In discussing the geology in the vicinity of Searsville, Whitney . says:* ‘*In the bed of the creek (one mile west of the ridge in which | the coal mine is situated) were, among the boulders of sandstone, some fragments of syenitic granite and of a basaltic rock, which latter is said to capa few of the highest points of this ridge.’ Whitney’s party also passed over the divide from San Mateo to Half- moon Bay, noting the Cretaceous and Miocene strata and the Mon- tara granite exposed along the road, but seem to have overlooked the diabase exposures on the east and south of the granite outcrop. _ W. L. Watts in a paper on San Mateo County says: ‘‘At some points basaltic rocks have been observed, and on the San Gregorio Rancho the Field Assistant of the Bureau noted and obtained specimens of vesicular dolerite, the vesicles of which were filled with petroleum.”’ 1 Published by permission of the Director of the U. S, Geological Survey. 2 The work of which the results are given in this paper was done while the authors were post-graduate students of geology in Stanford University, under the direction of Dr. J. C. Branner. The authors wish to acknowledge their indebt- edness to Dr, Branner for suggestions regarding the field relations, especially of the tuffs and associated rocks, and to ay: J. P. Smith for suggestions relating to the petrographical work, The names used in the lists of fossils in this paper are those commonly ap- plied to the respective species by the West Coast paleontologists. Owing to the imperfect state of our knowledge regarding the nomenclature of the California Tertiary fauna, there is a probability that some of the names used are erroneous: the writers, therefore, reserve the privilege of revising any or all of the names if fulure study shall warrant it, ‘Geological Survey of California, Vol. I, p. 71, 1865. * Tenth Ann, Rept, Calif, State Mineralogist, p. 586, 1890, 1904.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 17 Key Map FIELD RELATIONS. General Relations of the Diadbase., The exposures of the diabase have been traced for a distance of approximately thirty-two miles in a generally southeasterly direc- tion from a point near the San Mateo-Halfmoon Bay road, on Pil- arcitos Creek, in‘: San Mateo County, to a point on the head- waters of Lompico Creek, four miles east of the town of Boulder Creek, in Santa Cruz County. There is also a basaltic out- flow exposed near Stanford University, which is probably closely related genetically to the diabase. The exposures east and south of those found at the head of Devil’s Canyon are not shown on the map, as, with one exception, they are of minor importance. The largest exposed areas of the diabase are in the vicinity of Langley and Mindego Hills east of La Honda, and on the ridges between the headwaters of Pescadero Creek and the San Lorenzo River (the latter being outside of the area shown on the accompanying map). PROC. AMER. PHILOS. SOC. XLII. 175. B. PRINTED MARCH 9, 1904. te 18 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, The whole area presents at the surface a chain of more or less con- nected patches of diabase, extending approximately parallel to the coast, and also parallel to, but southwest of, the major axis of the Santa Cruz Mountains. The continuity of the patches is hidden by overlying strata and by dislocated masses of country rock and soil to such an extent that the exact relations of the various facies are difficult to ascertain. It is possible, however, to determine the age of the igneous mass by its relation to the sediments about it. The relations of the stratified rocks of the area under discussion are as follows: * Purisima formation with Astrodapsis n. sp., Lucina acutt- Pliocene | net, Nucula castrensis, Pecten 3 n. sp., A, B, and C, Rostellaria undulata and Saxidomus gibbosus. Monterey shale with Arca montereyana, Callista angus- tifrons, Pecten peckhami, and Tellina congesta. Vaquero sandstone with Agasoma barkerianum, A. kernianum, Pecten magnolia, and Turritella hoffmant. Miocene Associated Sedimentary Formations. The diabase proper breaks through beds of lower, and perhaps middle, Miocene age ; while the associated diabase tuff is interbed- ded with strata containing a typical lower Miocene fauna and lies below the Monterey shale. The basalt* outflow exposed near Stan- ford University overlies and metamorphoses beds of lower Miocene age, and is overlain by beds containing a fauna very similar to the underlying strata. This evidence indicates the lower Miocene age of the basalt and its probable contemporaneousness with the dia- base tuffs of Mindego and Langley Hills. Both the intrusive dia- base and the tuff are in many places overlain by the Purisima (lower Pliocene) beds, which show a distinct erosion line at their base, and also often a basal conglomerate made up of diabase peb- bles. 1 An uplifted mass of impure stratified limestone, containing a fauna that indi- cates its probable lower Eocene age, occurs in the diabase area between the headwaters of Pescadero Creek and the San Lorenzo River, This limestone appeirs to have no visible stratigraphic relations with the Miocene shales sur- rounding the diabase, * This basalt is the subject of a paper now in course of preparation by Prof. Milnor Roberts, of the University of Washington, id SS ee eee, Me 1904.] HAEHL AND ARNOLD—THE MIOCENE DIABASE. 19 Mouth of ark Canyon, on MiocEeNE.—(1) Zhe Vaquero Sand. a stone. The lower Miocene of the area consists of a series two or three thousand feet thick of mas- sive, coarse, yellowish sandstone layers, interbedded with a few layers of varying thickness of dark-colored argillaceous shale, the whole overlain by three or four hundred feet of thin- bedded siliceous shales. The lower part of this series of beds, including most of the sandstone, appears to have the same fauna and occupy the same stratigraphic position as the Vaquero sandstone of the Salinas Valley." The name ‘‘ Vaquero’’ will, therefore, be used to designate the lower Miocene sandstone in the area under discussion. The sandstone and shale series is typically developed in the region between the headwaters of Stevens Creek and the lower portion of Peters Creek. Fig. 1 shows a typi- cal section of this area. The fauna of the Vaquero sandstone series indicates its lower Miocene age. The following fossils, most of which are characteristic of the lower Miocene, are among others found in the Vaquero sandstone on Mindego Creek, Langley Creek, at the head of Stevens Creek, and at other points.in the area under discussion ; — “1r4S WL WIGVNT MMPS 1 ‘BL oanror Pd aw pera ves "SUOTJLULIOJ Jap[O pUk BUDIOIP 0} s¥quIP Jo sUONL[II SuLMoys *y9IID SUPAI}]S 0}U; aBplu wie Jaav puv YyaeID siajag da uorpes ysvayjiou-jsamyjnog 3N * The name for this sandstone was suggested by Mr. Homer Hanilin, and was first used by Dr, Fairbanks in the San Luis folio. This sandstone is typically developed in the Los Vaqueros Valley, near the Salinas Valley, Monterey County: 20 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, List of Vaquero (Lower Miocene) Fossils.—Those marked with a (*) are characteristic, so far as known.— *Agasoma barkerianum Cooper * Agasoma gravida Gabb * Agasoma kernianum Cooper * Arca microdonta Conrad Balanus estrellanus Conrad *Cardium (Trachycardium) n. sp. A. * Chione mathewsonii Gabb , Chione n. sp. A. (very large) * Conus n. sp. A. Crepidula grandis Midd. * Cuma biplicata Gabb * Dosinia conradi Gabb Dosinia mathewsonit Gabb * Dosinia cf. montana Conrad Dosinia aff. ponderosa Gray * Galeocerdo productus Agassiz , Galerus cf. excentricus Gabb *Glycymeris n. sp, A. (very large) * Lamna clavata Agassiz Lucina acutilineata Conrad Lucina richthofeni Gabb Mytilus mathewsonii Gabb Neverita recluziana Petit Ostrea titan Conrad Panopea generosa Gould Pecten estrellanus Conrad * Pecten (Chiamys) n. sp. E. *Pecten (Lyropecten) magnolia Conrad Pecten (Plagioctenium) n. sp. E. * Periploma n. sp. A. Psammobia edentula Gabb * Pyrula (?) sp. A. * Pinna alamedensis Yates * Siearetus scopulosus Conrad Solen sicarius Gould * Tivela inesiana Conrad * Trochita costellata Conrad * Turritella hoffmani Gabb * Turritelia ocoyana Conrad * Yoldia n. sp. aff. coopert Gabb (2) The Monterey Shale. The shales overlying the coarse yellow Vaquero sandstone are in some places thin-bedded, soft and chalky, while in others they are hard, dark colored and somewhat more massive. ‘The white facies of the shale is found overlying the diabase tuff in the region just west of the Langley Hill-Mindego Hill igneous area, while the dark colored facies is found on the northeast slope of the main ridge between the summit and Corte de Madera Creek. These shales represent at least a part of the Monterey series, which is supposed to be of middle Miocene age. ‘The upper part of the Vaquero sandstone series, at least that part showing alternating beds of sandstone and shale with a ten- dency to grade from the sandstone vertically upward into the shale, may be the inshore equivalent of some of the Monterey shale found at the typical locality in the region around Monterey. This 1904. | HAEHL AND ARNOLD—THE MIOCENE DIABASE. 21 theory is supported by the fact that, where typically developed, the Monterey shale is between twenty-five hundred and three thousand feet thick and rests on a comparatively thin layer of sandstone, while, in the area under discussion, the relative proportions of shale and sandstone are exactly the reverse. The harder, more flinty shales which appear along the coast are not found in the vicinity of the diabase. Fig. 2. View on the Searsville-La Honda road three-fourths mile south of sum- mit, looking east, showing the Miocene shale beds resting against the dia- base which has intruded them in sill-like dikes, The man points at the contact. Photograph by Ralph Arnold, The following species of fossils have been found in the Monterey shale within the diabase area: List of Fossils from the Monterey Shale (Miocene).—Those marked (*) are characteristic, so far as known.— * Arca montereyana Osmont * Leda sp. A. and B. Callista angustifrons Conrad Pecten peckhami Gabb Chione mathewsonit Gabb Pecten (Plagioctenium) n. sp. E. * Corbula sp. A. * Semele n. sp. A. Cylichna cf. petrosa Conrad Siliqua sp. A. Cythera cf. vespertina Conrad * Tellina congesta Conrad *Diplodonta n. sp. aff. serricata Reeve 33 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, aaa, ay ao a® 5 * a aaa a8 cs at a Mile Fig. 3. North-south section from Mindego Hill to Tar Creek, showing the relation of the diabase tuffs to the overlying Purisima beds. Puiocene.— Zhe Purisima Formation. Within the area under discussion is an extensive series of conglomerates, fine-grained sandstones and shales for which the writers propose the name ‘‘ Pur- isima formation.?’? This name has been chosen because of the typical development of the formation in the vicinity of Purisima Creek, San Mateo County. The Purisima beds lie unconformably upon the Vaquero sandstone and Monterey shale, and at the top grade into beds having a fauna somewhat similar to that of the Merced formation. Its upper limit may be defined as the base of the Merced. In age the Purisima probably represents the lower, and perhaps middle, Pliocene. The individuality of the fauna, stratigraphy and lithology of this formation appears to warrant the application of a new and distinctive name. Fig. 4 shows a typical section of the Purisima formation in the area. Meee aoe Purisirma shales and:3S..-800 feet. Nodular shalessss+rst? 200% ° Pliscene (Purisima) =) Green sandstone so’ ee | <4 Conglomerate + ++ +++:- RO « ol Lower and Middle Miocene 3500 Meecese Fig. 4. A typical section in the diabase area. Stratigraphically the Purisima formation presents a uniform cross 1 The junior author now has in course of preparation a paper giving in detail the characteristics and faunal relations of this formation, 1994.] HAEHL AND ARNOLD—THE MIOCENE DIABASE. 23 section throughout nearly the whole diabase area. At its base is the conglomerate, consisting of water-worn diabase pebbles ce- mented by a more or less siliceous sand. In some places, however, the amount of diabase is so great that it is difficult to distinguish the beds, where badly weathered, from the diabase tuff. The presence of barnacles (Ba/anus) and of a single specimen of Pecten afford the best evidence of the sedimentary nature of the deposit, and fix its origin as marine. The conglomerate is not always of this nature, however. South of the Alpine school- house the base of the Purisima consists of a rather incoherent shale breccia, which had its origin in a talus slope. Part of a Balanus was found in this breccia, showing that at least part of the deposit was laid down under water. In places fragments of the Miocene shale, together with hardened sandstone and chert, make up the greater portion of the basal layers, indicat- ing possibly that the Purisima coast line lapped over the intruded area and obtained its materials, not from the diabase area, but from beds to the east of them. The total thickness of the Puris- ima formation is probably about seven hundred feet. As a rule the conglomerate beds are thin; the thickest of them are about twenty feet in thickness. At some localities, notably at a place a quarter of a mile southwest of the Alpine schoolhouse, the base of the Purisima consists of shale, which rests unconformably upon the diabase. Above the conglomerate is a thin bed (four or five feet) of soft * green sandstone, stained by the chloritic weathering products of the diabase. It contains bones and sharks’ teeth and, in some localities, a rich marine invertebrate fauna. Over the green sand- stone is a bed of an unfossiliferous, nodular shale of perhaps two hundred feet in thickness. On the top of the unfossiliferous shale are sandy shales and fine sandstones probably five or six hundred feet thick. While these may readily be distinguished by their lithology, they are also char- acterized by numerous fossils which are in a fair state of preserva- tion. The following species were gathered from the Purisima beds in different parts of the area under discussion : List of Fossils from the Purisima (Pliocene) Formation.—Those marked with a (*) are characteristic, so far as known.— 24 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, * Arca canalis Conrad Mactra californica Conrad Arca trilineata Conrad Mactra falcata Gould * Astrodapsis n. sp. Merriam Modiolus rectus Conrad Astyris richthofent Gabb Mytilus mathewsonit Gabb Balanus estrelianus Conrad NVassa californiana Conrad * Buccinum n. sp. A. * Veptunea humerosa Gabb * Calliostoma n. sp. A. NVucula castrensis Hinds Callista angustifrons Conrad Olivella intorta Carpenter * Cancellaria n. sp. A. Otivella pedroana Conrad Cardium meekianum Gabb Panomya ampla Dall * Cardium meekianum n. var. A. Panopea generosa Gould Chrysodomus liratus Martyn Pecten expansus Dall Chrysodomus tabulatus Baird Pecten hastatus Sby. (smooth * Chrysodomus n. sp. aff. fabu- — var.) latus * Pecten n. sp. aff. expansus Crepidula grandis Midd. * Pecten n. sp. aff. dilleri Dall Crepidula rugosa Nuttall * Pecten n. sp. aff. parmeleet * Cryptomya n. sp. aff. californica * Priene oregonensis Redfield (n. * Dolichotoma n. sp. aff. carpen- var. P) teriana Purpura crispata Chemnitz Drillia incisa Carpenter Rostellaria indurata Conrad Galerus mammillaris Broderip Saxidomus gibbosus Gabb Glottidia albida Hinds Scutella interlineata Stimpson *ZLavicardium n. sp. aff. substri- Stliqua patula Dixon alum Solariella peramabilis Carpen- Leda cf. fossa Baird ter Leda taphria Dall Tapes staleyi Gabb Lucina acutilineata Conrad Tapes tenerrima Carpenter Lunatia lewisii Gould Tellina aff. congesta Conrad Macoma inquinata Deshayes Tresus nuttalli Conrad Macoma nasuta Conrad * Voluta n. sp. RELATION OF THE DIABASE TO THE ASSOCIATED SEDIMENTARIES, The masses of diabase follow the bedding, particularly in the shale of the lower Miocene series, and it is in between shale beds that most of the diabase exposures occur. Fig. 2 shows some shale beds resting against a large diabase dike which crosses the Sears- ville-La Honda road near the summit of the Santa Cruz range. The diabase dike in this case was intruded between and followed 1904.] HAEHL AND ARNOLD—THE MIOCENE DIABASE, 25 the bedding planes of the shale in the form of asheet. There are some very striking exceptions to the sheet-like occurrences of the diabase, but in general the ready cleavage of the shale along the bedding planes seems to have offered the line of weakness which the intrusive rock followed. Fig. 5 shows a characteristic case of the diabase breaking through the Miocene shales and sandstones. The shale in the middle of the dike in this exposure is slightly darker colored and somewhat harder than the shale beneath the diabase. ‘The sandstone was not affected by the intrusive rock. NE mV Ae eee SD s ! A pes SD N4 wer {kf} 4 #\r 6% ! INS Ses Ae ey oe Tey te PAA OMS Oe NINETY Col UN AG AVN AAAS ZOR Aan Ae, pe 7 02 NXD wie. 17 Diabase\e\i /i Diabase “2X8 XN ein Ca ol a Ao AS Sek ae ’s Se Shale Lyis Val Ny it ~ sey ae Pe, —} 4 <4" Pe a Sa alll A, “\ Liv ssl TAN WS Oy Fd dies Bere as Pot nin mee CO eet ath “: (Ht SY BES, OE ae Be ¥ OE Fig. 5. Vertical section exposed in small ravine on Dornberger’s ranch, near the Page Mill road summit, showing diabase intrusive in shale and between shale and sandstone. Inclusions of sandstone and shale are plentiful and vary from the size of a walnut to masses of hundreds of tons, but no evidence of the alteration of the sandstone has been noted, except in the case of the underlying beds of the basalt near Stanford University. Some well-preserved vertebrate bones and teeth (Oxyrhina tumu/la Agassiz) were found in a sandstone inclusion two feet in diameter about one-half mile north of the Alpine schoolhouse. Fig. 6 shows an inclusion of light yellow sandstone found near the edge of a large diabase dike on the Searsville-La Honda road. Neither the intruded sandstone, which is seen in the upper right-hand corner of the picture, nor the inclusion is in the least altered. The inclusions of shale are usually somewhat metamorphosed, but the metamorphism is not radical, changes in color and texture being the chief phenomena. An inclusion of shale four inches thick metamorphosed to a hard, brittle flint was found in the dia- base on Oil Creek. Similar occurrences were noted at several other localities in the area under discussion. Fig. 7 shows a shale layer which has been slightly hardened by an intrusion of diabase. This is an example of a somewhat common phenomenon. 26 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, Fig. 6. Inclusion of sandstone in diabase dike seen in vertical cut beside the Searsville-La Honda road three miles north of La Honda, Photograph by Ralph Arnold. * Fig. 7. View on the Page Mill road two hundred yards south of the summit, looking northeast, showing shale layer slightly hardened at the contact with the diabase dike, Photograph by Ralph Arnold, 1904. | HAEHL AND ARNOLD—THE MIOCENE DIABASE. 27 An interesting section (shown in Fig. 8) is exposed beside an old road one-half mile north of the Langley ranch house. The diabase at that place breaks through the Miocene shales, following the bed- ding planes in a general way, but sometimes breaking through the beds. Small inclusions of the shale are found in the diabase, but no alteration of either the inclusions or beds thus intruded was noticed. | RCE yenes eae SCALE West eee ' 2 5 4 Feet. Sha le —=——S Horizontal & Vertical Fig. 8. Vertical section of bank on south side of road one-half mile north of Langley’s ranch house, showing diabase intrusion in shale. The shale, which is unaltered, dips into the bank at an angle of 35°. The Purisima beds (Pliocene) which cover large areas of the dia- base are not penetrated by the diabase. This may explain the pres- ence of only small, isolated patches of the diabase along the north- ern end of the area. Either the Miocene or post-Miocene denu- dation over the northern end of the diabase area must have been great, or else the Monterey shale must be represented by sandstone over that territory, for there, wherever the diabase is exposed, it is in the Vaquero sandstones, while the Monterey shale appears to be almost lacking. Over this tract a deposition of the Purisima sedi- ments took place after the denudation and covered large areas of the eroded surface of the diabase. At the base of the Purisima beds are conglomerates made up largely of diabase pebbles, and these conglomerates are now exposed in the canyons together with small areas of the diabase in place. The presence of the diabase conglomerate at the base of the Purisima formation, together with the fact that the diabase is intrusive in the Miocene, establishes the time of at least the greater part of the igneous intrusions as later than the middle Miocene (Monterey), and before the Pliocene (Purisima). It is noticeable that the exposures of diabase in sev- 28 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb.5, eral instances are low down on the south-facing slopes of the ridges next to the creek beds, but are not visible on the north-facing slopes. This seems to be more than a mere coincidence. The dif- ference of exposure on the two sides of the canyon may be due partly to the thick vegetation and partly to the depth of decompo- sition and the admixture of organic matter in the formation of. the soil on the north-facing slopes. In general the Miocene shales near the diabase dip away from the intrusion as if it were the axis of an anticline. (See Fig. 2.) This may be due to a lifting action of the diabase upon nearly hori- zontal strata, or possibly to the fact that a pre-existing axis pre- sented the line of weakness along which the intrusion was made. There are instances of sill-like intrusions or sheets between the sedimentary beds of this area. The evidence of oil well records is available in some instances to show the presence of such sheets. © At well No. 1, on San Gregorio Creek near the mouth of Harring- ton Creek, the San Mateo Oil Company put down a test hole, and a sheet of diabase was encountered at a depth of one hundred feet. A hundred feet deeper the drill passed through the diabase and again entered the shale. This well is about a quarter of a mile from the igneous outcrop on Harrington Creek. Mr. Bell, on whose property another well was bored about ten years ago, is authority for the statement that no diabase was encountered in sinking that well, which is about four hundred yards west of San Mateo Oil Company’s well No. 1, and away from the diabase. ey ’ ° “ 1 mle Fig. 9. Northeast-southwest section through the Bella Vista oil well, San Mateo County, Photograph by Ralph Arnold. The well sunk by the Bella Vista Oil Company on the Bella Vista ranch, north of El Corte Madera Creek, encountered, according to the report of the driller, a fifty-foot stratum of diabase at a depth of four hundred and fifty feet ; the drill then passed into shale for another hundred feet, after which it again passed through diabase 1904.) HAEHL AND ARNOLD—THE MIOCENE DIABASE, 29 for fifty feet and again entered the shale. After passing through a few hundred feet of the shale the well entered the diabase again, and was still in the igneous rock when discontinued. Microscopic slides of the diabase encountered in this well showed it to be very similar to the exposure a quarter of a mile to the east, except that the rock was badly blackened with carbonaceous matter. The Tuff. The tuffs associated with the diabase are confined to the Langley Hill-Mindego Hill igneous area, of which they form the major por- tion. Within this area are also found diabase both of the diabasic and basaltic types, limestone beds, limestone dikes or intrusions, shale and sandstone. It is to be regretted that all of the rocks within this area cannot be differentiated on the map, as their areal distribution would throw much light on the structure of the terri- tory within which they occur. Beds of sandstone containing a typical lower Miocene fauna (given on a previous page) are found between layers of the tuff, while the shales containing Pecten peck- hami, when associated with the tuffs, are always found above them. This places most of the tuffs in the lower Miocene, with a possibil- ity of their extending into the middle Miocene. Layers of one of the basaltic facies of the diabase are found in such relation to the tuff as would indicate the contemporaneity of the two. This theory is strengthened by the fact that this characteristic basaltic facies, with the exception of the outflow near Stanford University, has been found so far only within the Langley Hill-Mindego Hill igne- ous area, to which the tuff is confined. The true diabase is later than the basaltic facies and associated tuffs, as it is intrusive both in the tuffs and in shale beds overlying them. The Purisima formation overlies unconformably both the tuffs and their overlying shale beds. (See Fig. 3.) The tuffs vary in composition from solid masses of basaltic dia- base fragments to almost pure limestone, sandstone and shale, de- pending on the conditions under which they were formed. It isa significant fact that the fragments of igneous rock in the tuff are, in all cases so far noticed, composed of the basaltic facies of the diabase. This is to be expected, as the extrusive forms of the rock would naturally be finer grained than the intrusive ones. The ma- terial in which the fragments of igneous rock are imbedded is gen- erally more or less limy, thus showing that the fragments were de- 30 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, posited in water at least deep enough to be the habitat of lime- forming organisms. ‘The theory that most of the tuffs were depos- ited in comparatively deep water is strengthened by the fact that the fragments in most of the beds are angular, which would not be the case had the tuffs been deposited near enough the surface of the water to be affected by the action of the waves. 2 * yg § 29 $4 996,02 0,5 1260. 34597 2 eS vals 79 itty AS REBSAR DA nO ehag Mag 50 2s’ r . latenwern : gu SSe—= ear Bae Fig. 10. Section exposed along the Searsville-La Honda road one-fourth mile north of La Honda, showing water-worn tuff interbedded between massive angular tuff. Fig. 10 shows a section which is exposed along the Searsville-La Honda road a quarter of a mile north of La Honda, Interbedded with the angular tuff is a layer of water-worn tuff about twenty feet thick. The angular tuff appears to have been deposited in the sea in successive layers until it reached near enough the surface of the water to be affected by the wearing action of the waves, when the water-worn layer was formed of the fragmental material. After a time a submergence took place and the top of the deposit was again lowered to such a depth as to be unaffected by the waves, or else the volcanic ejectamenta filled up the shallow sea quite above the water level. 4 ayaa aes * Sasale, nt 4@An , : , yr nA 18: i a ¢ r r to 0.48 “2: a) Cae eM 40.8 0 eG Gotta ay ys Arp warp Pr mae VS 8 het pm Pw aia Ad AydaB 7) a4 Verh PA P46 WAG wy OPAL RS 0 CET TNS ¢ ODD b G2 7 °*4 Dias ‘ © 4A £44 a esata aI ededgZl4y ae} LPTS ey rweayvé fs 6 n % Mile Fig. 11, Northast-southwest section through the top of Mindego Hill, showing the relations of the different tuffs exposed on it, 1904. | HAEHL AND ARNOLD—THE MIOCENE DIABASE, 3L Another interesting example of the relation between different facies of the tuff is shown in Fig. 11, which represents a section through the top of Mindego Hill. Here the angular tuff is over- lain by a water-worn tuff, which in turn grades by easy stages into a siliceous sandstone. ‘The line of demarkation between the angu- lar and water-worn tuffs is very distinct, the former being dark colored and grading into an almost massive basalt below, while the latter is composed of well-worn fragments of light-colored weath- ered amygdaloidal basaltic diabase. The water-worn layer grades into a tuff, which is composed of fragments of rock replaced by chalcedony, and then into a fossiliferous sandstone in which some of the fossils aad much of the rock have been replaced by chalce- dony. Chalcedony and quartz veins and chalcedony-lined cavities are common in the beds above the typical water-worn tuff. A peculiar tuff, composed of water-worn pebbles of the basaltic diabase imbedded in a fine, brown, ash-like matrix, is exposed on the Searsville-La Honda road just south of the mouth of Langley Creek. Where weathered this tuff so much resembles a true dia- base containing pebbles of basalt that at first its origin was quite Fig. 12. (@) Showing weathered and (4) fractured surface of the typical limy tuff from the hill north of the Langiey ranch house. Reduced one-half. Photograph by Ralph Arnold. 32 HAEHL AND ARNOLD—THE MIOCENE DIABASE. {Feb.5, puzzling. The layer is interbedded between shale layers and at first was thought to be intrusive in the shale, but a later examina- tion showed its true relation to the shales and its clastic origin. The typical tuff is found in thick beds all along the southwest- ern and part of the northeastern side of the Langley Hill-Mindego Hill ingneous area. Fig. 12a is a photograph of a hand-specimen showing a weathered surface of the tuff, while 124 shows a freshly fractured surface. The fragments composing the tuff are of dark- colored basaltic diabase, angular in outline and varying in size from the smallest grains to large masses weighing several hundred pounds. The slides of these fragments show them to be badly weathered, a few feldspars, a little augite and the magnetite and ilmenite being the only recognizable original constituents. The fragments are imbedded in a limy matrix, varying in composition from pure lime toa limy shale. Spheroidal weathering of the tuffs was noticed in one or two instances. Small organic remains are often found asso- ciated with the rock fragments in some of the more limy tuffs, Much, and sometimes all, of the lime occurs in a secondary form, as veins of calcite surrounding the fragments or cutting through the tuff. Pure calcite crystals weighing several ounces are some- times found in the tuff. This calcite is derived principally from the original lime beds in which the tuff was deposited, but a little Fig. 13. Thin section of diabase tuff, showing secondary calcite vein, (C); patches of secondary calcite, (C’); feldspar, (F) ; magnetite, (M). x 20. Photograph by Ralph Arnold, 1904.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 33 of it may come from a weathering of the feldspars of the basaltic fragments. Patches of isolated calcite are also found in most of the slides of both the basaltic and diabasic rocks. Fig. 13 shows a slide of diabase tuff from the hill east of the Langley ranch house. A secondary calcite vein (C) and small isolated patches of calcite (C’) are seen in this slide. Veins of chalcedony and limestone dikes or intrusions are also common in the tuffs. Limestone Dtkes. One of the most interesting phenomena met with in a study of the Langley Hill-Mindego Hill igneous area is the occurrence of limestone dikes or intrusions in the tuff beds. The best exposures of these dikes are found in the ridge to the north of the Langley ranch house. Figs. 14 and’ 15 show transverse and longitudinal sec- tions of this ridge, respectively. Similar dikes occur in the tuff 750 3 ‘am - moe <7 ive vu al “\7 Se NAS Fl kd, tat Severc ts ‘ aE Nal Nat os UNL a-lnm se ANICAUNIS7ESE t ms aden ve es ° 200 400° 600! Fig. 15. Northwest-southeast vertical section exposed on ridge north of the Langley ranch house, showing limestone layers interbedded with, and lime- stone dikes intrusive in, the diabase tuff. dred and fifty feet in length. Some of them show a kind of flow structure ; and a few of them show two systems of joint planes at right angles to each other and both perpendicular to the surfaces — et a > he ‘ ye ae py Lit * @ @ - 6. BF ie 5 “Ch Pd rp BP 4 yh oH 6 ining limestone dikes, found Fig 16, Vertical section showing detail of tuff conta Taken from sketch made on the ridge north of the Langley ranch house. in the field. 194.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 35 of the dikes. The surfaces of the dikes are quite irregular, giving a more or less wavy line in section, but the planes of contact of most of the dikes are approximately perpendicular to the bedding planes of the tuff and interbedded limy layers. Some of the dikes extend into the diabase which has intruded the tuff beds. Chalce- dony, quartz and calcite form veins and fill cavities all through the tuff, limy tuff beds and the limestone dikes. The minerals depos- ited from solution are of later origin than the limestone dikes. Fig. 16 shows in detail a small section of the tuff exposed on the side hill north of the Langley house. The chalcedony was deposited along a fault line developed after the intrusion of the limestone into the tuff. Fig. 17 is a photograph taken on the Searsville-La Honda road a quarter of a mile north of La Honda, and shows the tuff cut by limestone dikes and calcite veins. Fig 17. Vertical section along the Searsville-La Honda road one fourth mile north of La Honda, showing limestone dikes (D) and secondary calcite veins (V) in the diabase tuff. Photograph by Ralph Arnold, The origin of the limestone dikes is easily accounted for when the relations of the containing and associated terranes is consid- 36 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, ered. The series of beds in which the dikes occur north of the Langley house have an upward sequence of sandstone, tuff, limy shales, and then alternating thick beds of tuff, comparatively thin beds of limestone and limy tuff, the whole capped by sandy tuff, above which are shale and sandstone beds (see Fig. 14.) Soon after the deposition of this series, and before the tuffs and lime- stones had become very coherent, diabase was intruded between the lower sandstone layer and the overlying tuffs. The intruded bed fractured the tuff along lines approximately perpendicular to the bedding planes. of the series, and the unconsolidated ooze and limy tuff of the interbedded layers flowed into the fissures, thus forming the dikes. The Diabase. Studied in the field the diabase presents two facies. One will be termed the diabasic, the other the basaltic. The distinction is made purely on the physical appearance of the two. No great chemical difference exists, but the crystallization, color, texture and manner of weathering are so radically different that, while no dif- ferentiation is attempted on the map, a distinction is necessary in describing the rocks microscopically. Secondary dikes of small proportions were found in the diabasic type, and will be briefly described under that head. The diabasic facies.—The diabasic type seems to be confined to the masses which make the north and east boundaries of the area between the south fork of Tunitas Creek on the north and Langley Hill on the south, In all cases it lies along the crest of the range, making the highest peaks and giving them a peculiar rounded out- line that is readily distinguishable at a distance. The rock is well exposed near the summit of the ridge, on the road which crosses the range two and one-half miles south of Sierra Morena, Here the course of the dike is plainly marked by the large rounded boulders on the hillsides. The rocks weather in such a way here as to give particular prominence to the feldspars, thus giving the mass the appearance of a gabbro. The soil derived from its disin- tegration closely resembles granitic soil. It is made up of granular particles with a slight reddish cast, and varying in size from a diameter of one quarter to one-sixteenth of an inch. Macroscopically the rock is a medium grained, light gray, crys- talline aggregate, in which three components are very readily dis- 1904.] HAEHL AND ARNOLD—THE MIOCENE DIABASE, 37 tinguishable. One, augite, is present in dark patches intruded by the others, and showing distinct glistening cleavages. Magnetite can be detected in large flat plates and smaller grains, dark and lustrous. Separated with a knife-blade, small portions can be picked up readily with a magnet. The most evident component is the feldspar. It occurs in long white rod-like crystals sometimes as much as two inches in length, giving a reticulated appearance to the mass; it is banded and contains inclusions of magnetite and augite. Fig. 184 is a photograph of the typical diabasic facies, being specimen No. 24, the analysis of which is given as I in a following paragraph. Fresh specimens showing but slight kaolinization are readily ob- tained. Occasionally a crystal is seen to contain a few clear, glassy spots quite easily distinguishable with a hand lens. They are prob- ably analcite. eee hai talts Se, ER” Fig. 18. (@) Showing the basaltic (specimen 38) and (4) true diabasic (speci- men 24) facies of the diabase. Reduced one-half. Photograph by Ralph Arnold, On the stage road from Redwood City to La Honda, at a point on the west side of the summit, about one-half mile from the Summit 38 HAEHL AND ARNOLD—THE MIOCENE DIABASE. _ [Feb.5, House, the road cuts across the contact of the diabase with the lower Miocene sediments in a small canyon, so that a good cross section is exposed (see Fig. 2). The sediments here dip at an angle of sixty degrees south, twenty degrees west, and the diabase, which is of the coarse variety and rather badly weathered, follows the bed- ding planes. Within thediabase, running exactly parallel to the contact and dipping with the sediments, are a number of coarsely crystalline secondary dikes, varying in width from one inch to six inches and standing out hard and fresh in the darker decomposed diabase. Figure 2 is a photograph of this dike. The secondary dikes may be seen to the left of the man in the figure. Macroscopically this rock is medium grained, light colored and with a rather mottled appearance, due to the uneven distribution of the more basic constituents. Augite, plagioclase and magnetite are the chief components. The augites are large and tend to segregate in spots, often with a poikilitic structure; the feldspars being in- cluded in the augites and giving a mottled appearance to the rock. The feldspars are long and narrow, somewhat kaolinized and show banding. Clear patches of analcite are frequently included in them. Magnetite is very plentiful in long irregular blades which stand out prominently and often reach a length of half an inch. A few small cavities in the rocks are filled with a mass of rathér flexible, fine, acicular crystals matted togéther indiscriminately. The crystals are usually light colored, although a few are discolored,. evidently by weathering products. Such small amounts were ob- tained that it was impossible to determine them accurately. Before the blowpipe they are infusible and they are not acted upon by acids. The basaltic facies. —It is thought best to treat all the fine- grained dark varieties which make up the remaining portion of the area under the head of basaltic facies. These in turn could be read- ily separated into at least two general types, differing in the coarse- ness of their crystallization and weathering products, but such a classification would be tedious and will not be attempted. It is necessary to state, however, that those portions of the area which are made up of smaller dikes are almost universally of a coarser texture than the larger masses and exhibit spheroidal weathering in a very striking way. Figure rg illustrates a typical example of the spheroidal weathering of the medium grained diabase. The basaltic facies differ from the diabasic facies in that they are dark, show- 1904] HAEHL AND ARNOLD—‘HE MIOCENE DIABASE. 39 Fig. 19. Spheroidal weathering of the diabase exposed beside the Page Mill road, one-half mile east of the summit. X 5. Photo. by R. Arnold, ing the white feldspars but indistinctly, the predominating crys- tals being augites and olivines. The finer grained varieties make up the larger masses, such as the tuffs and some of the dikes of the Langley Hill-Mindego Hill igneous area, and are often amygda- loidal, weathering into a compact adobe soil. Amygdaloidal cav- ities of great size are frequently encountered. One cavity filled with quartz measured four inches along its greatest diameter. Cal- cite, chalcedony and serpentine fill the cavities in many in- stances, and on Bogess and Harrington Creeks diabase in place was found with its vesicles filled with petroleum. Perhaps the most interesting occurrence is the presence in many places of nests of glassy analcite crystals, filling the amygdaloidal cavities and joints and seams in the rock. Almost perfect icositetrahedrols were obtained. Qualitative tests showed the presence of Al, Naand SiO,. The mineral is fusible before the blowpipe and soluble in hydrochloric acid, yielding no jelly, however; in this particular agreeing with the observation made by Lawson and Palache’ on 1¢The Berkeley Hills,” Bui. Dept. Geol. Univ. Cul., Vol. II, p. 418, Berkeley, 1902. 40 HAEHL AND ARNOLD—THE MIOCENE DIABASE, | [Feb. 5, analcite from the andesites in the Berkeley Hills. At several points in the area small irregular aggregates, varying in size from one-tenth to one-half of an inch in diameter, made up of fan-shaped growths of slightly clouded, white crystals, were found in the weathered dia- base. When tested before the blowpipe these crystals were found to be natrolite, and a thin section cut from one of the small bodies showed the angular centre area between the natrolite crystals to be filled with analcite (see Fig. 25). Calcite veins of considerable size were found in the mass in sume places. Macroscopically the rock is fine grained and dark. Augite and olivine crystal are readily detected in most specimens, sometimes in crystals large enough to be porphyritic. Plagioclase feldspar and magnetite are also present, and pyrite has been found in a few places. The augite is dark and lustrous and usually quite fresh. The olivine, however, is generally somewhat weathered to serpentine, which often fills the crystal cavity completely and gives the rock a greenish tint. Another weathering product of the olivine was found quite plentifully in thin scales of light brown color. Chem- ical tests showed the presence of Na, Ca, Fe and Mg. The mineral is hydrous and infusible. Treated with hydrochloric acid, it becomes lighter in color and gives up its iron. These, together with its op- tical properties, which will be mentioned, make it possible that it is the mineral described by Lawson' as iddingsite. The feldspars in this facies are almost universally microlitic. An occasional pheno- cryst is seen. Magnetite is present in small grains, barely visible to the unaided eye. The basaltic facies usually has a very distinct, coarse, conchoidal fracture. Figure 18a is a photograph of speci- men No. 38, a piece of the typical basaltic facies, the analysis of which is given as II ona following page. Microscopic PETROGRAPHY. The petrographic discussion contained herein is based upon the study of about one hundred and thirty slides, cut from the rocks col- lected over the diabase area and examined under the microscope. In thin section the eruptive presents two facies. Both are holo- crystalline and contain about the same minerals, but the one pre- sents a rather granular structure under the microscope, just as it does in the hand specimen; the other a finely crystalline, aphanitic '“The Geology of Carmelo Bay,” Bull, Dept. Geol, Univ, Cal., Vol. I, p. 31. 1904.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 41 structure with smaller individuals, yet of nearly the same chemical composition. While no separation of the two has been attempted in mapping in the field, they will be treated under separate heads in dealing with their microscopic character. In describing their field relations the first has been called the diabasic, the latter the basaltic facies, It is not intended that these terms shall be used to desig- nate two distinct series of rocks, but rather in the sense of a con- venient classification of two facies of the same magma. Fig. 20. Thin section of the diabase facies (specimen 24), showing the typical diabase structure. (A), augite; (F), feldspar; (M), magnetite. x 20. Photograph by Ralph Arnold. The Diabasic Facies. Considering the general tendency of the eruptive to disintegrate, the diabasic type is usually remarkably fresh and clear in thin sec- tions. The slides show the following principal constituents, given in the order of their crystallization: magnetite, ilmenite, apatite, olivine, feldspar, augite and analcite. The last is never present as an original constituent, so far as could be determined, but is cer- tainly in many cases, and probably in all cases, a secondary pro- duct. Of the secondary minerals, serpentine, chlorite, iron ores, calcite and natrolite have been noted. Plagioclase.—The feldspar is generally present in the diabasic facies in rather stout, lath-shaped forms with an average length of two millimeters, twinned according to the Albite and Carlsbad laws 42 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb.5, Fig. 21. Twinned crystal of labradorite, showing cleavage. 60. (see Fig. 21). Of the two, the Albite twinning predominates and is usually polysynthetic, and occasionally combined with the Per- icline. Crystals cut parallel to the composition plane and showing tabular forms are not infrequent, and in many cases show zonal structure and wavy extinction, indicating a centre more basic than the periphery. Extinction angles were carefully taken and indicate plagioclases of about the order of labradorite with a formula of mixture about Ab,An,. Decomposition ot the feldspars has gone on to a great extent in portions of the mass. Comparatively fresh sections are obtainable, however, in places. Kaolinization is very common in all sections. The most characteristic alteration, however, seems to be that which results in the formation of analcite within the feldspar. Nor does it seem that any one law of decomposition applies to all the cases seen. In one instance a mere patch of an isotropic, clear glassy min- eral is found in the centre of a plagioclase. In another the crystal form of the feldspar appears to be filled with the product, except, perhaps, a small fresh patch of the original mineral left in the cen- tre, in just such a position as the analcite held in the first case cited. Occasionally the whole crystal is replaced by the analcite. In the slides examined there seems to be much evidence that the analcite is an alteration product (in these instances) of the feldspars them- selves. The problem of the percentage of soda required for the 1904.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 43 formation of analcite will be discussed under the head of ‘*Chemical Characters.’’ Lenses of high powers reveal the presence in the feldspars of many dust-like particles, the nature of which is unknown. Usually they are without definite arrangement. Inclusions of gas bubbles, patches of augite, magnetite, and also serpentine and chlorite are noted. Augite.—Augite is very plentiful in the diabasic facies of the rock, and is usually allotriomorphic with respect to the feldspars. It is of a pale brown cclor with a tinge of red, probably due to a small per- centage of titanium—a supposition which the rock analysis appears to verify. Its pleochroism is faint, changing the shade and not the color. Extinction angles as high as 53° were noted, and zoned crys- tals with undulatory extinction were occasionally seen. Twinning is not uncommon. Cleavage cracks are very distinct, and the inter- secting cleavage lines parallel to the prism of 87° 6’ are frequently observed. ‘The augite is remarkably fresh and clear in this rock, having withstood the effects of weathering better than the feldspars. Smaller crystals of augite, occasionally included in the larger phen- ocrysts, are often almost entirely decomposed into what appears to be a yellowish-brown chlorite, the coloration being due to the iron ores present. Frequent irregular patches of gas and fluid inclusions occur in the phenocrysts, sometimes long and rope-like, and often clustered around smaller included grains of augite. Irregular in- clusions of feldspar are often found and are generally much kaolin- ized. Magnetite and its decomposition products are also present in the phenocrysts. Otivine.—Olivine is not abundant in the slides of the diabasic facies. It would have been possible, however, to so choose the sec- tions as to show considerable of this mineral, as its occurrence seems to be in occasional local patches and segregations. It is pres- ent, however, in very small quantity in the typical slides, usually in minute clear patches, making up the centre of a mass of brownish decomposition material, badly discolored by iron and showing no characteristic optical properties. Its crystal form, where dis- integration is complete, suggests its origin from olivine. In rare instances, too, this secondary decomposition mass assumes a fibrous structure, strong pleochroism and strong double refraction with bright red and green polarization colors, suggesting iddingsite.' 1«The Berkeley Hills,” by A. C. Lawson and Chas, Palache, Budi. Dept Geol. Univ. Cal., Vol. II, No. 12, p. 430, Berkeley, 1902. ons HAEHL AND ARNOLD-—THE MIOCENE DIABASE. _ [Feb.5, Other Minerals.—Magnetite and some ilmenite are present in the diabase as grains and irregular masses. Grouping is occasionally seen, but for the most part both of these minerals are scattered through the mass without definite position. Apatite is sparingly present in its usual characteristic clear, long, slender prisms, included by the other constituents. Of the secondary products analcite is by far the most important. It occurs chiefly in the feldspars in the diabasic facies, In no in- stance could it be shown that it is other than a secondary product, nor does it indicate an origin other than of an alteration product of the feldspars. Treated with hydrochloric acid it is soluble, but forms no jelly. Fig. 22, Section of secondary dike. (A), analcite; (P), augite; (F), feld- Spar. X 30, The Secondary Dikes.—In thin sections the rocks of the second- ary dikes contain apatite, magnetite, augite, sphene, feldspar, pyrite, analcite and natrolites (see Fig. 22). The sections are particu- larly clear and fresh, The feldspars are plagioclases with the composition of labrador- ite. They are broadly lath-shaped and show Albite twinning. Wavy extinction with a basic interior and more acid periphery is common 194.) HAEHL AND ARNOLD—THE MIOCENE DIABASE. 45 Kaolinization is somewhat advanced; dust-like inclusions, together with augite and magnetite, ate frequent. The augites are of the pale purplish-brown variety with slight ple- ochroism. They make up an unusually small percentage of the rock, however. Basalsections show wavy extinction. Both idio- morphic and allotriomorphic crystals are present. Cleavage is distinct and relief high. Inclusions of feldspar and magnetite are numerous and decomposition very slight. Magnetite is present in unusual quantities and in very striking, long, slender rods, as well as in its common tabular forms. A few crystals of pyrite were noted, alsoa wedge-shaped crystal of sphene. Analcite and natrolite are present in these sections in greater quantity than in those of any other portion of the mass. Fig. 23. Section of the basaltic facies (specimen 38), showing basaltic and flow structure. (QO), olivine; (0/), olivine weathering to iddingsite; (F), feld- spar crystal with etched edges; (A), augite. x 20. Photograph by Ralph Arnold, The Basaltie Factes. In thin section the basaltic facies of the igneous mass presents a more difficult problem than the diabasic type, because it is univer- sally more weathered. The typical section shows a few phenocrysts of olivine and augite in a fine-grained ground mass of lath-shaped feldspars, microscopic augites and olivines, ilmenite, magnetite, and the secondary products—calcite, serpentine, chlorite, iddings- ite, iron oxides, natrolite and analcite. 46 HAEHL AND ARNOLD—THE MIOCENE DIABASE, _ [Feb.5, feldspar.—The feldspars are in two general forms—typical lath- shaped crystals and broader tabular plates with wavy extinction or zonal structure and, usually, an abundance of inclusions. The lath- shaped crystals predominate. As far as it was possible to deter- mine them, both seem to be of the order of labradorite and exhibit the same extinction angles that were characteristic of the plagio- clases of the diabasic type. Individual crystals are seldom over one millimetre in length. Twinning is usually polysynthetic ac- cording to the Albite law, although Carlsbad twins are frequently noted. Flow structure is often beautifully shown by the arrange- ments of the lath-shaped microlites in regular courses between the larger crystals of augite or olivine (see Fig. 23). Not infrequently the ophitic structure of the typical diabase is seen, the feldspars ra- diating, in all cases noted, around the larger crystals of olivine. This is particularly true of the slides cut from the rocks of the narrow dikes. In numerous cases, especially in sections showing flow structure, feldspars are bent and broken and displaced. In many slides the ground mass is badly decomposed and shows practically no optical phenomena, except such as is shown by the feldspar mi- crolites which, in these sections, are so badly etched along the crystallographic outlines that they present rough, saw-like edges (see F, Fig. 23). Feldspars differing from the general type are oc- casionally found in the slides. Slides from one exposure show feld- spars which at first glance might be mistaken for orthoclases, so clear and regular are they and free from banding or twinning. In- clusions are numerous, however, and a closer examination of their optical properties leaves little doubt that they belong to the plagio- clases. There is an unusual amount of analcite in these slides, which suggests very strongly that the feldspars may contain a larger percentage of soda. Again some sections show feldspars, the order of whose interference colors borders closely on nepheline, and one slide shows a number of crystals whose optical properties would tend to class them as melelites. In view of these resem- blances, tests were made upon the thin sections to determine phys- ically whether the optical properties were true indicators. The re- sults, however, left no doubt that the crystals were simply feldspar. Crystals were also found in these slides, portions of which gave the normal optical phenomena of the feldspars common to the rock. Low order interference colors are frequently met with in the more weathered slides, but in nocase could nepheline be positively detected. 194.) HAEHL AND ARNOLD—THE MIOCENE DIABASE, 47 The matter of the presence or absence of the nepheline was made the object of particularly careful search, as its presence, if estab- lished, would materially assist in accounting for the soda necessary to the formation of analcite, as has been observed by Fairbanks in dealing with a very similar rock in San Luis Obispo County. As shown above, however, it is very doubtful whether any nepheline occurs in either facies of the diabase of the area here under discus- sion. Pyroxene.—The pyroxenic constituent is usually augite, but en- statite is occasionally noted. The augite occurs, in general, in two ages—a more or less porphyritic series which are occasionally idio- morphic and frequently absent entirely in the slides, and a series of small allotriomorphic grains filling the interstices between the feld- spars and olivines of the ground mass. Augites of this latter type are seldom over five-hundredths of one millimetre in diameter, and seem to be identical in composition and optical phenomena with the porphyritic type. No grouping ofeither type could be detected, the only instance of a perceptible order of arrangement being found in the slides from one small area on Harrington Creek, where por- phyritic augites with distinct micropoikilitic structure were ob- served. The included crystals were particularly fresh plagioclases, - which made up about fifty per cent. of the cross section of the py- roxene host. The phenocrysts seldomattained a large size in this facies and were usually broken by mechanical strains or rounded and etched by chemical action, However, elongated crystals with ap- proximately idiomorphic outlines were not uncommon. The augite is of the same pale brown to pinkish tint noted in the diabasic facies. Like it, too, it is but slightly pleochroic, except for some few scattered individuals whose pleochroism is somewhat marked. Po- larization colors are very brilliant. Twinning according to the aug- ite law is not uncommon. Only simple twins were noted. Inclu- sions of glass, gas bubbles and magnetite were noted in the porphy- ritic crystals. Enstatite is found in a few instances in irregular plates showing low interference colors. The crystals were in no case large, two- tenths of one millimetre being the greatest diameter measured. The surfaces of the crystals were distinctly pitted, but no distinct 1«Analcite Diabase,” by H. W. Fairbanks, Bu//, Dept. Geol, Univ. Cal., Vol. I, pp. 273-300, Berkeley,*1895. 48 HAEHL AND ARNOLD—THE MIOCENE DIABASE. _ [Feb. 5, cleavage was observed. Irregular patches of the enstatite were found included in the feldspars. On the whole the pyroxenes are remarkably fresh. Many slides show absolutely no decomposition products even where the feld- spars are badly weathered. In a few instances slight chloritization was noted, that being practically the only indication of decomposi- tion. . Olivine.—The olivine, like the augite, is present in two genera- tions, porphyritic and microlitic. The porphyritic crystals are usu- ally idiomorphic and are among the oldest segregations of the magma. They are usually much fractured and jointed, and rounded or embayed by the corrosive action of the magma. Usually disin- tegration has gone on to such an extent that only the crystal form remains, filled with the secondary products. Where the original olivine remains it is clear and colorless, with strong double refrac- tion and high relief. It usually includes much magnetite in small grains, some glass and dust particles. ee = tgs \k Ruts oot Rt atte 4e ; ¢ \ : . x "* : “4 ~_~ - ae * a >. | > + tet f as et . °. AA e ‘ ee i te ? wi 7 nf \ 7 re = ’ Dh oe) e aS” os AG +\ el Poy Ae fo °0 * + omen fs BACs ete Oe Pa vor et a Bos Fig. 24. Phenocryst of olivine (O) weathering to iddingsite (1). x 60. The most common product of decomposition is serpentine, which frequently shows its fibrous character. Alteration begins along the cracks, gradually working inward from them until the crystal is di- vided into a number of irregular rounded patches of clear olivine separated by fibrous serpentine, the whole making up the complete form of the original crystal. Perhaps fifty per cent. of the pheno- erysts studied were completely replaced in this way, the remainder ° showing various stages of such decomposition or alteration in like manner to the mineral, which is probably iddingsite (see Fig. 24 1904.] HAEHL AND ARNOLD—THE MIOCENE DIABASE, 49 and O', Fig. 23). It shows high polarization colors and is strongly pleochroic in the green shades, the greatest absorption being paral- lel to the fibres. It agrees strongly with the mineral described by Iddings' from Nevada and afterwards named iddingsite by Lawson. A small amount of chlorite was noted. Calcite is quite abundant in the more weathered portions of the rock. It is usually found filling seams, joints and amygdular cavities. Fig. 25. Natrolite (N) and analcite (A) between crossed nicols. XX 30. Anailcite.—The most striking product of decomposition is anal- cite. Its occurrence in the field has been described. In certain areas it is quite plentiful. In thin section it is often found associ- ated with natrolite, fibrous aggregates of which it frequently in- cludes (see Fig. 25). It is isotropic, occasionally showing optical anomalies. It has been observed to occur in five general ways: 1. In irregular patches in the centre of crystals of plagioclase. 2. In a form suggesting a decomposition product of the plagioclases, ad- vancing in irregular lines from the crystal edges inward. 3. Com- pletely filling what seems to have been the rectangular outline of a plagioclase crystal. 4. In irregular patches filling the angular 1« Geology of the Eureka District, Nevada,” by Arnold Hague, Mon, XX, U, S. G. S., Appendix B, pp. 388-390, Washington, 1892. PROC. AMER. PHILOS, SOC. XLIII. 175. D. PRINTED APRIL 2, 1904. 50 HAEHL AND ARNOLD—THE MIOCENE DIABASE. [Feb. 5, Spaces between crystals of feldspar. 5. In large irregular patches sometimes two and a half millimetres in diameter, filled with a con- fusion of microlites of some indeterminable mineral. CHEMICAL CHARACTERS AND ANALYSES. The writers are indebted to the United States Geological Survey for the analyses (I and II) of the two typical facies. For the pur- pose of comparison and discussion there have been added analyses of the analcite from Cuyamas* (III), the teschenite diabase (IV), plagioclase feldspar (VI) and analcite (VII) from Point Sal,’ and a typical analysis of labradorite from Dana (V).° ‘6 Il. III. Ty; Vs VI. VII. SiO, ..\s sever 50.12 49.60 50.55 49.61 56.0 52.72 54.40 ALO, .. x os sieeaiee 18.52 16,56 20.48 19.18 27.5 30.46 23.04 Fe,O, . + deueee 2.47 4.28 2.66 212 O07 FeO |... Bapee 4.11 4.44 4.02 5.01 MgO ..... saiaoe 2.68 5.38 4.24 4.94 O.I CaO >. ie ) AF CQibils Gayo FO 59 12s were purle- “up tn ahoult Soyears” Ar aprascla err & annum = 13,000 000 cya remared from bed —e re 1982. == : ss aay a Value at sects per ey fh 40, 000, ore. ul) 7433 ALQUTAE WMOTES. ‘The Tarmpo crevasse hevosi¥etg 4/3 <9 qls, in 30 years . or 3£00g dye Cy. Cac year, which, @ (2 fer and for lire 3° years, wourachk mace The costs g oS coco These lwo crevasses déschare abou 18% of fee sill-an- mually carrie £6 She Gulf, Wilhoue a 4 he i, . vavigable channel wo d r \ ins seed 5 QW Port Eads SOUTH PASS Oerra of THE Scale, Sas” making about one-sixth of the sediment which is thus withdrawn by these crevasses, with consequent benefit to navigation, reduction of flood height and increase of the public domain, all without cost.* It would therefore seem that these /atera/ outlets have very much 1 The depth of three yards was obtained by taking the average of a number of soundings covering the areas filled up by the crevasses (8.7 ft.). 94 HAUPT—THE MISSISSIPPI RIVER PROBLEM. [Feb. 19, to commend them to the consideration not only of engineers but of economists, business men, farmers and real estate dealers; and that, so far as the evidence of nature goes, their operation is only bene- ficial and in no wise injurious. If they were closed and the river leveed, all of the advantages named would be destroyed and the sediment would be carried to the mouth, where it would extend the bars more rapidly, raise the flood plain and require elevation of the entire system of levees along the river banks. But there is another class of outlets which may be considered in this connection, and that is the bars which obstruct the mouths and thus prevent the free discharge of the fluvial waters. These may be distinguished as /ongitudinal outlets, and their permanent removal is entirely practicable by applying the energy of the river to the. work to be done. It is a well known fact that a sedimentary stream, flowing through a straight reach, seldom maintains a single permanent channel, while in swinging around curves the concave bank, acting as the directrix, causes a reaction which deepens the bed and deposits the silt upon the complementary convex bank which is the resultant of this action. z In this way, by the operation of natural laws, the deposits are removed from the path of navigation and the cross-section is auto- matically adjusted to the requirements of the river. Instead, there- fore, of building two parallel jetties as substitutes for the natural banks, and thus extending the river into the Gulf at the expense of its slope andthe reduction of its area of discharge along straight lines, which are unnatural and unfavorable, it will be found more rational to build one curved training wall so placed as to create a head and reaction which will transport the silt to the opposite or convex bank, where it will be deposited without cost, leaving an ample navigable channel and saving the expense of one of the jetties, while it also scours away the bar directly in front of the mouth and affords an open passage for the effluent water. By thus utilizing the tendency of water to flow in curved lines instead of straight ones half the cost of the jetty works may be saved and a better and more permanent channel be obtained, with a low- ering of the flood heights of the river. This result is due to the form of the orifice, and it will be seen that when no such modifica- tion is applied the effluent stream is abruptly checked by the inertia of the Gulf water and the sediment thus deposited acts as a xe 1904.] HAUPT—THE MISSISSIPPI RIVER PROBLEM. 95 buffer, to divide and deflect the energy into lateral components, which are again subdivided indefinitely, as shown in the typical forms of the deposits at Cubit’s Gap; whereas when supported and concentrated by the continuous reaction of a properly placed resist- ing medium, the activity of the currents thus generated will prevent deposits near the trace of the work and create a neutral zone or counterscarp at some distance therefrom, which will thus become the site for the dump. These features may be observed wherever there are obstacles placed in the path of a current. The best arti- ficial illustration of the efficacy of this principle as applied to a tidal inlet with a feeble tide is to be found at Aransas Pass, Tex. The lowering of the flood plain by the removal of the barriers to the longitudinal discharge is also well illustrated by the operations on the River Tyne in England, where the flood heights have been reduced from nine to three feet along the stream by the opening of the mouth and removal of the bars from the bed. Regardless, therefore, of the interests of navigation, it would be of great benefit to the State and nation to open the mouths of all the Passes for drainage and reclamation purposes, and by the use of the proper form of tool this could be accomplished more effectively and at less cost than by the methods now in vogue at the mouths of sedimentary rivers. In the foregoing analysis it has been the intention to lay particu- lar stress upon the necessity of so regulating the movements of the sediment as to prevent its being deposited in the pathway of the stream, where it may operate to obstruct its flow, causing elevation of bed, banks, levees and greater risks and expense; for it is evi- dent that so long as the commingled earth and water are confined to the channel with no avenues of escape, the deposits must engorge the bed and involve continuous danger and expense. It would seem that the attention of the engineering profession has been focussed mainly upon the control of the water, apart from its sediment, and with secondary consideration to the evils result- ing from failure to separate these two elements, which, it is believed, may be done to great advantage at a number of points ex route where lands may be reclaimed by the natural process of hydraulic grading, and large tracts of the richest arable land be recovered in a comparatively short time at a cost which will be insignificant as compared with that required to grade and drain it by mechanical means. 96 HAUPT—THE MISSISSIPPI RIVER PROBLEM. [Feb. 19, 1904. It is therefore desired to direct particular attention to the neces- sity of providing suitable dumping sites for the mud carried sea- ward by the river in times of flood, where it may be deposited beyond the banks of the stream without injury, through or over suitable weirs, and be retained by impounding dikes in the low swampy regions to their advantage. The question is similar in its general features and effects to that prevailing at the inlets along alluvial coasts, where it was the practice to attempt the removal of the bars by jetties in pairs, sup- plemented by dredging ; but which method has not been able to meet fully the demands of modern vessels, so that recourse is now being had to the control of the heavier earthy materials which compose these obstructions, in such manner as to protect the channels from their encroachments and cause the single concave jetty to construct and maintain much greater depths than exist in a state of nature. Stated Meeting, March 18, 1904. President Smiru in the Chair. A letter was read from the Marquis de Nadaillae accepting his appointment as the Society’s representative at the cele- bration of the Centenary of the Société Nationale des Antiquaires de France. The decease was announced of William Marriott Canby, of Wilmington, Del., on March 10, 1904, et. 73. Prof. Felix E. Schelling read a paper on “ The Academic Drama in the Age of Elizabeth and James.” at Sah oy PROCEEDINGS AMERICAN PHILOSOPHICAL SOCIETY HELD AT PHILADELPHIA FOR PROMOTING USEFUL KNOWLEDGE Vou, XLII. AprIL, 1904. No. 176. General Meeting, April 7, 8 and 9, 1904. Aprit 7.—Mornina@ Session, 10 A.M. President Smrrx in the Chair. The President opened the General Meahing with a brief Address of Welcome. A letter was received from the Committee of Organiza- tion of the Fourteenth International Congress of Americanists, to be held in Stuttgart, August 18 to 23, 1904, inviting the Society to be represented at the Congress by a delegate; and Also from the International Zoological Congress, to be held at Berne, August 14 to 19, 1904, inviting the ROCIeLY to be represented by delegates at the Congress. On motion, the President was requested to appoint delegates to these Congresses. From the President of the American Academy of Political and Social Science, inviting the members to attend the sessions of the Academy on April 8 and 9. The following’ papers were read: “Dimethyl Racemic Acid, Its Synthesis and Derivatives,’’ by Prof. H. F. Keller, of Philadelphia. Discussed by Prof. George F. Barker. “The Réle of Carbon,” by Prof. Albert B. Prescott, of Ann Arbor, Mich. PROC. AMER. PHILOS. SOC. XLII. 176. @. PRINTED MAY 18, 1904. 98 MINUTES. [April 7, 8 and 9, “Sources of Error in the Determination of the Atomic Weight of Nitrogen,’ by Prof. Theodore W. Richards, of Cambridge, Mass. Discussed by the President. “The Constituents of the Venom of the Rattlesnake,” by Prof. John Marshall, of Philadelphia. Discussed by Mr. Joseph Willcox, Prof. William B. Scott and Dr. Marshall. “Trisulphoxyarsenic Acid,” by Prof. LeRoy W. McCay, of Princeton. “The Atomic Weight of Tungsten,” by Prof. Edgar F. Smith and Mr. F. F. Exner, of Philadelphia. Discussed by Prof. Barker and Prof. E. J. Houston. “The Expansion of Algebraic Functions at Singular Points,” by Prof. Preston A. Lambert, of Bethlehem, Pa. Introduced by Prof. C L. Doolittle. AFTERNOON Session, 2 P.M. Vice-President Scorr in the Chair. “The Continuum and the Theory of Masses,” by Prof. I. J. Schwatt. Introduced by Prof. C. L. Doolittle. “An Attempt to Correlate the Marine with the Fresh and Brackish Water Mesozoic Formations of the Middle West,’ by Prof. John B. Hatcher, of Pittsburg, Pa. Discussed by Profs. Osborn, Scott and Heilprin. “The Miocene Rodentia of Patagonia,” by Prof. William B. Seott, of Princeton, N. J. Discussed by Prof. Osborn. “Recent Advances in Our Knowledge of the’ Evolution of the Horse,” by Prof. Henry F. Osborn, of New York. Dis- cussed by Profs. Conklin, Heilprin, Scott and Mr. Willcox. “The Yukaghir Language,’ by Mr. Waldemar Jochelson, of New York. Introduced by Dr. Franz Boas. Discussed by Dr. Franz Boas. “The Horizontal Plane of the Skull,” by Dr, Franz Boas, of New York. tn) ares 1904,] MINUTES. 99 “The Silurian Fauna of Arkansas,’ by Mr. Gilbert van Ingen. Introduced by Prof. W. B. Scott. “Palladium,” by Mr. Joseph Wharton, of Philadelphia. Discussed by Mr. Willcox and Prof. Baskerville. EvEeNING Session, 8 P.M. |(At the Free Museum of Science and Art, University of Pennsylvania.) President Smiru in the Chair. The following paper was read: “Pompeii and Saint Pierre: An Examination of the Plinian Narration, and Other Studies” (with lantern slide illustra- tions), by Prof. Angelo Heilprin, of Philadelphia. Fripay, Aprit 8.—Mornine Session, 10 A.M. Vice-President BARKER in the Chair. The following papers were read : “The Reflex Zenith Tube,” by Prof. Charles L. Doolittle, of Philadelphia. Discussed by Prof. Snyder and Dr. Brashear. “Faint Double Stars,’ by Mr. Eric Doolittle, of Philadel- phia. “On the Spectra and General Nature of Temporary Stars,” by Prof. William W. Campbell, of Mt. Hamilton, Cal. Dis- cussed by Dr. Brashear and Dr. Barker. “Systems of n Periplegmatic Orbits,” by Prof. Edgar Odell Lovett, of Princeton. Introduced by Prof. C. L. Doolittle. “Radium from American Ores,” by Prof. A. H. Phillips, of Princeton, N. J. Discussed by Mr. Joseph Wharton and Profs. Barker and Phillips. 100 _ MINUTES. [April 7, 8 and 9, EXECUTIVE SESSION, 12.15 P.M, President Smiru in the Chair. The report of the Committee appointed to prepare a plan for the appropriate celebration of the bi-centennial of the birth of Benjamin Franklin was presented, and on motion the Com- mittee was continued with power to carry out the plan. The pending nominations for membership were read and the Society proceeded to an election. AFTERNOON Session, 2 P.M. President Smrrx in the Chair. The Tellers reported that the following candidates had been elected to membership: Residents of the United States— Maurice Bloomfield, Ph.D., LL.D., Baltimore. Henry Pickering Bowditch, M.D., LL.D., Se.D., Jamaica Plains, Mass. Edward Potts Cheyney, A.M., Philadelphia. Russell H. Chittenden, Ph.D., New Haven. Frank Wigglesworth Clarke, 8.B., Se.D., Washington. John Chalmers DaCosta, M.D., Philadelphia. Kuno Francke, Ph.D., Cambridge, Mass. Adolphus W. Greely, U.S.A., Washington. Preston Albert Lambert, Bethlehem, Pa, Edgar Odell Lovett, Ph.D., LL.D., Princeton. Edward Leamington Nichols, Ph.D., Ithaca, Hon. Theodore Roosevelt, Washington. Samuel W. Stratton, Washington. Harvey W. Wiley, A.M., M.D., LL.D., Washington. Foreign residents— Friedrich Delitazsch, Berlin, Sir Richard C, Jebb, Cambridge. “Seay 1904.] MINUTES. 101 Ernest Rutherford, Montreal. Jakob Heinrich Van’t Hoff, Berlin. Wilhelm Waldeyer, Berlin. The following papers were read: “A System of Passenger Car Ventilation,” by Dr. Charles B. Dudley, of Altoona, Pa. Discussed by Mr. Willcox, Dr. Marshall and Prof. Heilprin. “ Atmospheric Nucleation,” by Prof. Carl Barus, of Provi- dence, R. I. Discussed by Profs. Kraemer, Conklin and Snyder. “On the Classification of Meteorites,’ by Dr. Aristides Brezina, of Vienna. “Doliolum and Salpa,” by Prof. William Keith Brooks, of Baltimore. Discussed by Prof. Conklin. “On the Breeding Habits of the Spade-Foot Toad (Scaphi- opus solitarius),” by Dr. Charles Conrad Abbott, of Trenton. N. J. “On the Occurrence of Artifacts Beneath a Deposit of Clay,” by Dr. Charles Conrad Abbott, of Trenton, N. J. “The Organization of the Germ Cells and Its Bearings on Evolution,” by Prof. Edwin Grant Conklin, of Philadelphia. “The Origin and Nature of Color in Plants,” by Prof. Henry Kraemer, of Philadelphia. Saturpay, Aprin 9.—MorninG Session, 10 A.M. President SmirH in the Chair. “The Establishment of Game Refuges in the United States Forest Reserves,” by Mr. Alden Sampson, of Haverford, Pa. Discussed by Profs. Morse and Hewett, Dr. Brashear and the President. “The Use of the Relative Pronouns in Standard English Writers,” by Prof. Waterman T. Hewett, of Ithaca, N. Y. 102 PRESCOTT—THE ROLE OF CARBON. [April 7, Discussed by Dr. Brashear, Prof. Morse, Mr. Yarnall and Prof. Hewett. “The Effect of the American Revolution Upon the English Colonial System,” by Mr. Sydney George Fisher, of Philadel- phia. Discussed by Mr. Stuart Wood. “The Hedonic Postulate,” by Prof. Lindley M. Keasbey, of Bryn Mawr, Pa. Discussed by Mr. Stuart Wood, Prof. Doolittle, Mr. Richard Wood and Prof. Keasbey. “Results of the American Ethnographical Survey,” by Prof. Marion D. Learned, of Philadelphia. Discussed by Mr. Rosengarten, Mr. Richard Wood and Mr. R. P. Field. “Regulation of Color-Signals in Marine and Naval Ser- vice,” by Dr. Charles A. Oliver, of Philadelphia. “The Ripening of Thoughts in Common,” by Prof. Otis T. Mason, of Washington. THE ROLE OF CARBON. BY ALBERT B. PRESCOTT. (Read April 7, 1904.) It may be said of any one of the chemical elements that it acts a part of its own in the formation of matter and the manifestation of energy in the world. A chemical element taken as it is, aside from questions of its genesis and its decay, stands out before exact meas- urement as an innate individual factor in the production of things throughout the universe. Whatever there is now being brought to light between matter and the ether or the electrons, at all events the chemical elements taken in their atomic quantities are the present facts upon which further inquiry must rest its advances. The behavior of an element is an experimental constant, however progressive may be the theories by means of which men of science may pursue their studies. The present is for some reasons a time profitable for us to recount certain of the salient characteristics of that chemical element named at the head of this brief paper. The registration of carbon compounds in M. M. Richter’s Lexi- con, amounting to 80,000 in the year 1900 and since increased by RI 1904.] PRESCOTT—THE ROLE OF CARBON. 103 * the addition of two supplements, presents for our consideration something besides the choice of man in the direction of his chemi- cal research. These advancing thousands of individual combina- tions, of determinate molecular weight and fixed elemental compo- sition, give us evidence of the chemical productivity of carbon and of its character in relation to the other elements. We need to keep before us the place of carbon among its rela- tives in the periodic system. Central as it is in its electric polarity and in the order of its valence, the leading member of a group holding an equilibrium among the other groups, its place is that of a balance of power. But an element, preéminently this element, is more than the occupant of a place, more than a mere number in a progressive system, a mere function of a weight; it is all of these perhaps, but if soit is more: it is an individual. Carbon is not wholly exceptional, however, neither in the sense of an entire lack of the variability of neighboring elements, nor in that of being the only one whose atoms can at all unite to each other in the forma- tion of chains. It is by virtue of both its central position and its independent character that it appears, when in combination, as the element in command. We must recognize the fact, without explanation, that the unusual ability of carbon to unite its atoms in chains is dependent upon their union with hydrogen, whose aid is thereby indispensable to the organic world. Carbon is formed for complexity only when supported by the unvarying-unity of hydrogen, and by this support is provided the great capacity of carbon for extensive molecular structure. The science of chemistry, and therefrom all physical science, has been enriched by experimental studies of molecular constitution. These are studies of determinate facts, and it is but a consistent expression of these facts that is undertaken in structural formulz or even in the atomic theory itself, as used in the work of chemists. The several differences, for instance, between dimethyl ether and ethyl alcohol, two individuals of the composition C,H,O, are differ- ences of fact. We give statement in the figurative language of the structural formula and of the atomic theory to the actual nature of the one as an ether and of the other as an alcohol. It is a discov- ered truth that in the alcohol one-sixth of the hydrogen is united to the oxygen more intimately than the other five-sixths of the hydrogen are united to the oxygen. It is a truth that in the ether 104 PRESCOTT—THE ROLE OF CARBON. [April 7, the union of each sixth of the hydrogen to the unit of oxygen is the same. Were chemists to abandon structural formule, were they to go further and desert the atomic theory, all the facts heretofore communicated would remain to be told if possible in other terms to the eye and ear. There has been some reaction against the devotion paid to mole- cular constitution, and there may well be a protest against certain besetting tendencies in the teaching of this subject. Both teacher and investigator ought to be on guard against the assertiveness of the structural formula. Let us welcome as a corrective the in- creasing service of empirical formulz, the appearance of the formula index once a year in the journals, the constant uses of the Lexicon of Carbon Compounds, and the adoption of empirical formulz frequently for summation and comparison as well as for contrast. Of course I refer only to these formule when of determined mole- cular weight, and we must recognize that no small share of the vantage ground in organic chemistry since 1890 is indebted to the new methods of molecular weight determination. It is the nature of the carbon atom that has made attractive to chemists the work they have done upon molecular structure. It was long ago established that.the character of a compound depends partly upon what elements unite to make it and partly upon the order of their union. No atom in a molecule is wholly without influence upon every other however remote. And as to the effect of any element in a compound, it is a fair conclusion that the num- bers of its atoms within the molecule count for something, the rela- tive position of its atoms may count for more, the structural concentration of its atoms will count for most. When the nature of the proteids and other matters manifesting life shall become known we may be sure that molecular constitution will be included in that knowledge. We may be sure that the carbon atom, or some theoretical equivalent of what we now term the atom, will be a very determinate part of the question. This is not to say that chemical synthesis alone can compass vitality, but rather that vitality must still depend upon the chemical constitution of the vitalized material. We cannot speak lightly of the limits of what may come to be defined as the molecule, or count confi- dently upon future restrictions of its extent. We may well admit, however, that the rim of the molecule wherever placed must continue to bound the province of chemical study, Ann Arbor, Mich., March 317, 1904. 1904.] KELLER AND MAAS—DIMETHYL RACEMIC ACID. 105 DIMETHYL RACEMIC ACID: ITS SYNTHESIS AND DERIVATIVES. BY HARRY F. KELLER AND PHILIP MAAS. (Read April 7, 1904.) In the course of an investigation on diacetyl, the simplest diket- one of the aliphatic series, Fittig’ and one of us obtained a crystal- lized acid of the composition C,H, ,O, + H,O, which they recog- nized as a dimethyl derivative of racemic acid. Continuing the work on diacetyl, we prepared the diketone on a larger scale, and it occurred to us to utilize the accumulated residues for a more extended study of dimethyl racemic acid. In view of the close relationship and the striking resemblance which this compound bears to the tartaric acids, and of certain interesting questions which had been suggested in the original research, it seemed well worth while to take up this line of work. Unfortunately the time we could devote to it has been very limited, so that the results we have to present are somewhat fragmentary, and much remains to be done. The conversion of diacetyl into the desired acid was effected by the same process as that employed by Strecker in his well-known synthesis of racemic acid, that is by successively treating the diket- one with hydrocyanic and hydrochloric acids. Thus OH / Crete 0 HON- | CH. con ! — 1 CH,—C—0+ HCN CH,—C-CN N\ - OH OH OH H,—CHcNn H,—c. + 40+ Ha ey aN Ch CHyo¢= cK CH, —¢— COOH IS mh OH OH These reactions, consisting in the addition of hydrocyanic acid to an aldehyde or ketone, followed by ‘‘ saponification ’’ or hydro- lysis of the resulting cyanhydrin, are regarded as generally appli- cable to aldehydes and ketones and are, of course, familiar to every 1 Liebig’s dnnalen der Chemie, Vol. 249, p. 208, 106 KELLER AND MAAS—DIMETHYL RACEMIC ACID. [April 7, student of organic chemistry. Nevertheless only meagre descrip- _ tions of the manner in which they have been applied in the various cases are to be found in the special literature of the subject, and no little difficulty was experienced in ascertaining the conditions favorable to the formation of the dicyanhydrin of diacetyl and its conversion into dimethyl racemic acid. Numerous and varied were the attempts to effect a quantitative union of the diketone with hydrocyanic acid, but in no case did the yield of the dicyanhydrin exceed 30% of the theoretical. Large quantities of other products, among them the monocyan- hydrin, were always formed. The method generally recommended, that of Wislicenus and Urech,! in which the compound containing the ketonic group, CO, is made to react with hydrocyanic acid . in the nascent state, gave very poor results, and the only way in which larger quantities of the desired product could be obtained, was that, originally used by Fittig and Keller. It consists in mixing diacetyl with aqueous hydrocyanic acid. The best results were obtained by adhering to the following direc- tions: The diketone in portions of about 20 grammes is gradually added to a 30% solution of hydrocyanic acid, the heat of the reaction being checked by cooling with water. The mixture is then placed into a pressure bottle and heated in the water bath at about 60° for some hours. The product of the reaction may now be extracted with ether (in which it is quite soluble), or it may be directly converted into dimethyl racemic acid by treating the solu- tion with hydrochloric acid. The former seems preferable, since it facilitates the purification of the final product. The saponification of the crystallized dicyanhydrin presents no difficulties. Recent experiments have shown that it proceeds rapidly and nearly quantitatively, when the substance is heated with ordinary hydrochloric acid under pressure at 100°. After remov- ing the excess of hydrochloric acid by evaporation on the water bath, the residue, containing much ammonium chloride and some tarry matter, is dissolved in water, the solution filtered, and the filtrate carefully neutralized with baryta water. It is then boiled with further additions of baryta (to decompose the ammonium chloride), and allowed to stand after having. been made slightly acid with acetic acid; barium dimethyl racemate separates as a 1 Liebig's Annalen der Chemie, Vol, 164, p. 255. ae ve 1904.] KELLER AND MAAS—DIMETHYL RACEMIC ACID. 107 characteristic crystalline precipitate. Its separation may be pro- moted by the introduction of a few particles of the solid salt. The free acid is readily obtained from this barium salt by means of sulphuric acid. The theoretical quantity of the latter, diluted with water, is added to the finely powdered substance, and the mixture heated with occasional stirring. When the decomposition appears complete, the barium sulphate formed is filtered off, and the filtered solution evaporated to crystallization. The dimethyl racemic acid thus prepared generally forms large transparent crys- tals, and is easily purified by repeated crystallizations. This pro- cess of extracting the dimethyl racemic acid from the products of the action of hydrochloric acid and diacetyl dicyanhydrin, is more convenient and gives a better yield than that described by Fittig and Keller. More than 60 grammes of material have been made by its means. Properties of Dimethyl Racemic Acid.—The physical properties of the acid have been accurately described by its discoverers, and little can be added here to their description. Although remarkably fine crystals have been repeatedly obtained, a definite determina- .tion of their form remains to be made. There is reason, however, for believing that racemic acid is ”o¢ isomorphous with its dimethyl derivative. Contrary also to a previous statement, it has been observed that when crystals of dimethyl racemic acid are kept for a long time they do effloresce, though far more slowly than those of racemic acid. The water of crystallization is completely expelled at 105°, rede- terminations of its amount gave 9.05% and 9.12 %; theory requires 9.18%. Above 110° there is a further loss of weight, owing no doubt to a slow decomposition of the molecule. The solution of dimethyl racemic acid in water, like that of other synthetic compounds containing assymetric carbon atoms, is opti- cally inactive. Salts of Dimethyl Racemic Acid.—With the very limited amount of acid at their disposal, Fittig and Keller were able to prepare but a small number of the salts, and three of these only were obtained in a pure state and in quantities sufficient to permit analytical determinations. ‘The data thus secured were, however, sufficient to confirm the composition of their acid, and to establish its close analogy with the tartaric acids. At the same time certain anom- alies in the composition and some of the properties of the dim- 108 KELLER AND MAAS—DIMETHYL RACEMIC ACID. [April 7, ethyl racemates were observed, which rendered a more elaborate study of these salts desirable. This was the object in the experi- ments recorded on the following pages. The neutral salts of sodium, potassium and ammonium, which are very soluble in water, were made by neutralizing weighed portions of the acid with calculated quantities of the alkaline carbonates or ammonia, and evaporating to crystallization. Sodium dimethyl racemate, C,H,O,Na,-+ 4H,O, forms minute, efflorescent needles. A freshly prepared specimen yielded 23.85% of water, while two others which had been exposed to the air for some time, gave 17.65% and 16.14% respectively. Theory re- quires 24.48%. The amount of sodium in the anhydrous salt was— Required. Found. TRG aes aa Saas s «0's crates kh aoe 20.72% 20.75 % Both the meufra/ and the highly characteristic acid potassium sait have been previously described. A new determination of the potassium contents of the latter gave 17.87%, instead of 18.05% required. The ‘‘ Rochelle Salt ’’ of dimethyl racemic acid appears to have the composition C,H,O,KNa + 2H,O. Its preparation and analy- sis were attended with many difficulties. On neutralizing the acid potassium salt with sodium carbonate and carefully evaporating the solution, a homogeneous crop of crystals was obtained, but on further evaporation the liquid deposited a mixture of efflorescent prisms and wart-like aggregations of clear anhydrous crystals. An analysis of the first product yielded— Requires, Found. De shine ce anek + hacdicesu™ caeen anil cendeel314% 4 12.00% K (calculated for anhydrous compound) ..... 16,25 “ 16,12 Na “ “ Oa wees 9.58« 9.49 * Unlike the salts of the alkali metals, those of the alkaline earth and heavy metals are insoluble or nearly insoluble in water, though in a few instances they are difficult to precipitate. Many of the insoluble salts contain water of crystallization, which they tena- ciously retain at high temperatures. In’ some cases the. composi- tion corresponds to that of the racemate, in others there are notable differences both in composition and properties. The calcium salt which, owing to its great insolubility, was recommended by Fittig and Keller as the most delicate means of 1904.) KELLER AND MAAS—DIMETHYL RACEMIC ACID. 109 recognizing the presence of dimethyl racemic acid, is at once pre- cipitated when calcium acetate or gypsum solution is added to that of one of the soluble salts, It may be obtained in a crystalline form by dissolving it in hydrochloric acid and reprecipitating with sodium acetate. Its composition varies considerably according to the conditions under which it is formed. The water of crystalliza- tion is not completely expelled below 215°, a temperature just below incipient decomposition. The following are a few of the analyses made of this salt: Required. Found. 1mol.H,O0 1%mol.H,O I. Th: Ee FSD sin 3 iis a 7:65 % IL.11% 7:38% .... 11.39% ct vdeis eda Ss SE 16.46 “ 17.20% 16,12 16,33 * The dartum salt, CH,O, Ba + 2 H,O, shows a remarkable re- semblance to the corresponding racemate, from which it differs, however, in that it contains one-half molecule less of water of crystallization. This point has been established by comparative analyses of both compounds. ‘The results were— oe Racemate. — Dimethyl racemate, Required. Found. Required. Found. H,O ....13.64% 13.52% 10.32% 10.12% 10.42% Ba..... 41.40“ 41.10 “ 39-25 “ 39-14 39.16 « Magnesium salt, C(H,O,Mg + H,O. A white bulky precipitate. _ The air-dried substance gave— Required. Found. ERP AEM DA y sia hapa eh s.n's a> 0's\dme.c ceeeee 8.25% 7.06 % PSM N Agr) cleiid 1Gda eK bAe 9a oc pda oN IL,01 “ 10.95 Manganese salt, CS3H,O,Mn + %4H,0. Forms a crystalline precipitate having a faint pink tint. The formula is derived from these results : Calculated. Found, RENTS WARN ee is's Sa ese enc ss nedseaews s 3-75 % 3-72% UTE Maa Pedic Katies 615.5 sign o é pies cheelaae 22.91 * 22.97 « The zinc salt, C,H,O,Zn + H,O, is obtained as a granular, white precipitate when zinc acetate is added to an alkaline salt of dimethyl racemic acid. It is almost insoluble in water and in dilute acetic acid. Determinations of the water of crystallization 110 KELLER AND MAAS—DIMETHYL RACEMIC ACID. [April 7, in different preparations of the salt indicate that it can crystallize with either one-half mol. or one mol. of water. One specimen yielded 3.76% and 3.55% H,O (3.60% corresponding to 14 mol.), while the analysis of another gave these results : Calculated. Found. RRIOALA EROSIN NSS we 6.95 % 6.87 % CA aati seh Hague tion 25.18 « 25.26 « Several attempts to estimate the zinc electrolytically gave lower results, it being found impossible to effect a complete deposition of the metal from a cyanide solution of the salt. Cobalt and nickel salts. A curious difference was repeatedly noticed in the way in which these compounds separated from the solutions in which they were formed. The cobalt salt is invari- ably precipitated in the cold, and nearly quantitatively, from the liquid, while the solution in which the nickel compound is formed remains perfectly clear for days or weeks even, until it is heated or evaporated ; a precipitate or residue is then formed which is about as insoluble as the cobalt compound. Both salts deposit as crystal- line crusts ; the color of the cobalt salt is a fine purple, that of the nickel salt apple-green. Both contain one molecule of water of crystallization, and their compositions correspond to the formule C,H,O,Co + H,O and C,H,O,Ni ++ H,O. Calculated, Found. FIO cocsn pwr cvviene cece 7.11% 6.13% and 6.74% NAc n05 i tina opeseeeeees’s 23.72 23.09 ** 23.34% BUD a she > cbeeon-e nove eee THOR 6.73 * 6.85 « Buea aN Sees ce vers sb eneeees 23.23 °¢ 23.19 22.76 «6 The metals were determined both as sulphates and electrolytically. Cadmium salt. This is very similar to the zinc salt, but differs from the latter in being anhydrous. Its weight remains constant to above 160°. Its solution in potassium cyanide was electrolyzed without any difficulty, and the electrolytic determination was checked by one in which the cadmium was weighed as oxide. Required, Found. 2 Il, CA Vellicke Galen evita pG 38.61% 39.20% 38.78 % Lead salt, C(A,O,Pb +- 2H,O. In the earlier part of this inves- tigation, considerable quantities of this compound were obtained as 1904.] KELLER AND MAAS—DIMETHYL RACEMIC ACID. 111 an intermediate product in the preparation of pure dimethyl racemic acid. Its formation and properties have been described by Fittig and Keller, who were unable, however, to spare any of their material for analysis. A carefully prepared specimen afforded the following determi- nations : Required. Found. BEAD ia lilads ich EARL. Rabe Khe d deed ne 8.59% 8.52% EA pa dice tasesigawae'e corcccccecese + «49.40% 49.226 The copper sa/t is anhydrous. When a solution of copper acetate is added to a soluble dimethyl racemate, the liquid remains per- fectly clear, but on acidifying it with acetic acid, a light green amorphous precipitate is produced. It is probable, therefore, that this copper compound when first formed exists in the colloidal state. The weight of the air-dry salt remained constant on heating; it yielded— Calculated. Found, MDE Cae oe oa 2 ie seis ume oe Vales 26.39 % 26.49 % Silver sait, A voluminous and amorphous precipitate, which becomes denser on standing and darkens when exposed to light, results upon addition of silver nitrate to a neutral solution of the sodium salt. It is likewise anhydrous. Calculated. Found. Wa 35 canes 61 ea bee 8 eb 55-10% 54.76% and 54.92% When we compare the above results with the composition and properties of the corresponding tartrates, and more particularly the racemates, we are forced to admit that the acid obtained from dia- cetyl bears the closest resemblance to the group of isomeric com- pounds, of which it is the dimethyl derivative. Such differences as have been observed between dimethyl racemates and racemates are no greater (if indeed as great) than those known to exist among the salts of the several modifications of tartaric acid. Action of Heat upon Dimethyl Racemic Acid.—It was noticed by Fittig and Keller that when the acid is heated to 178°-179° it melts with partiai decomposition, and is completely volatilized, without charring, when the temperature is raised sufficiently. _ Experiments, which have thus far been made on a small scale only, indicate the presence of at least one, and probably of two 112 MCCAY—TRISULPHOXYARSENIC ACID. [April 7, acids among the volatile decomposition products. In one case a few grammes of the acid, contained ina miniature retort, were slowly heated in the paraffin bath, the volatile products being con- densed in well-cooled tubes. ‘Two distinct stages in the decompo- sition were observed. Just above the melting point the substance began to boil, giving off pungent-smelling vapors. The aqueous distillate had a strongly acid reaction, and on extracting it with ether and evaporating this solvent, a small quantity of needle-like crystals, having acid properties, was obtained. When the residue in the retort was heated to about 250°, renewed boiling, and more vigorous than before, took place. The distillate, which was of course separately collected, was oily and viscous, and was found to consist largely of an ether-soluble acid. It has not been prepared in sufficient quantity to permit an analysis. The ammonium salt crystallizes in little prisms, and its solution yields precipitates with calcium, barium, lead and silver salts. The lead salt deposits in beautiful stellated aggregates and seems very characteristic. The study of the decomposition products of dimethyl racemic acid will be continued as soon as enough of the starting material can be procured. Central High School, Philadelphia, April 7, 1904. TRISULPHOXYARSENIC ACID. BY LEROY W. MCCAY. (Read April 7, 1904.) About eighteen years ago, in order to account for the irregu- larities accompanying the interaction of sulphuretted hydrogen and arsenic acid, I assumed the existence of three sulphoxyarsenic acids lying between arsenic and sulpharsenic acid : H,As0, H,AsO,S H,AsO,S, H,AsOS, H,AsS, That the monosulphoxyarsenic acid can exist in the free state I showed in 1886. A few months later Preis, of the University of isle iiss NS teeta By NEN hechieee aes Mun 1904.] McCAY—TRISULPHOXYARSENIC ACID. 113 Prag, established the existence of the disulphoxyarsenic acid, and recently Dr. William Foster, Jr., and I have succeeded in preparing several salts of the trisulphoxyarsenic acid. Ever since Preis dis- covered the disulphoxyarsenic acid I have been convinced that the trisulphoxy-compound existed, and since it was the only acid necessary to complete the series, a great deal of experimental work was undertaken in order to isolate it. . Dr. Foster’ published recently the results of an elaborate investi- ; gation of the action of magnesium oxide on a mixture of equivalent amounts of arsenic trisulphide and sulphur suspended in water, it being hoped that during the reaction the magnesium salt of the acid would be formed in sufficient amounts to make possible its transfor- mation into the corresponding sodium salt, and the separation of this by means of alcohol. His work has served to clear up many matters bearing on the modes of formation of the sulphoxy- compounds; and although no perfectly consistent results were reached, sodium salts were prepared again and again whose compo- sition approached so closely to one corresponding to the formula Na, AsOS, +-11H,O that I felt convinced the substances were really impure tertiary sodium trisulphoxyarseniate. A few preliminary experiments having established the fact that freshly precipitated arsenic pentasulphide suspended in water is decomposed far more readily by an excess of magnesium oxide than a mixture of equivalent amounts of arsenic trisulphide and sulphur, I suggested to Dr, Foster that we conjointly make a careful exam- ination of the resulting solution. The proposal was a profitable one, for we find that the solution contains large quantities of mag- nesium trisulphoxyarseniate. : The magnesium salt of trisulphoxyarsenic acid, then, is formed when magnesium oxide in excess acts upon freshly prepared arsenic pentasulphide suspended in water kept during the entire reaction at o° C. The change is rather slow, when the amount of arsenic pentasulphide is large, but it is generally complete in four to five hours. - By precipitating the magnesium in solution by means of sodium hydroxide, as magnesium hydroxide, adding an equal volume of alcohol to the filtrate from the magnesium oxide and hydroxide, and keeping the corked flask and its contents in the ice chest, the 1 Z. anorg. Chem., 37, 64 (1903). PROC. AMER. PHILOS. SOC. XLII. 176. H. PRINTED MAY 19, 1904. é 114 McCAY—TRISULPHOXYARSENIC ACID. [April 7, tertiary sodium salt of the acid commences to separate out in feathery crystals which, in the course of twelve hours, pass over | into fine fern-like forms. The compound is purified by recrystalli- zation. The yields are very satisfactory. Thirty grams of As,S; yield about thirty grams of the impure salt. The compound possesses a composition represented by the formula Na,AsOS, ++ 1111,0. Calculated. Found. PIO Cease ce eee hse cee tee erechs 15.22% 15.65% PEGs eee ian 5 Cae Ore 16,50 * 16.84 * SER wae Riis Samedihs + dau bebaegeisids 3-526 3-45 “ SSE a EL ES pS Fe er 21.166 20.74 * 1 9 BRET Po re, MY pee 43-60 * 43-32 100,00 100,00 The tertiary potassium salt is prepared in an analogous manner. On the addition of the alcohol it separates out in the form ofa light yellow oil, which, however, solidifies to a straw-colored, crystalline mass when kept for some hours at —20° C. It crystallizes with seven molecules of water : K,AsOS, ++ 7H,0. By adding an alcoholic solution of strontium chloride to an aqueous solution of the sodium salt, the double trisulphoxyarseniate of sodium and strontium is precipitated in a crystalline condition : NaSrAsOS, + 10H,0. Barium chloride produces in a solution of the tertiary potassium salt a crystalline precipitate of potassium barium trisulphoxy- arseniate : KBaAsOS, -+- 7H,0. Aqueous solutions of sodium trisulphoxyarseniate are not precipi- tated by strontium chloride. This reaction has been made use of for separating the small amount of monosulphoxy-salt which is occasionally thrown down along with the trisulphoxy-compound. Barium chloride precipitates both the di- and trisulphoxyarsenic acids, but the barium salt of the latter acid is more soluble than that of the former. The behavior of these two acids toward hydrochloric acid is also a means of distinguishing between them. 1904.] McCAY—TRISULPHOXYARSENIC ACID. 115 If a dilute solution of sodium trisulphoxyarseniate be treated with enough acid to render it strongly acid, then shaken violently and filtered, the filtrate is clear and becomes but faintly turbid on boiling. If a dilute solution of sodium disulphoxyarseniate of the same concentration be tested in a similar way, the filtrate becomes strongly turbid on boiling. When these tests are made in flasks and the flasks, immediately after the addition of the hydrochloric acid, are stoppered, so as to prevent the escape of the sulphuretted hydrogen, at the end of thirty-six hours no smell of the gas can be detected in the flask which contained the disulphoxy-salt, while it is still very pronouncéd in the one which contained the trisulphoxy- compound. ‘These two reactions have been studied very carefully. It appears that the trisulphoxy-acid breaks down as follows : 2H,AsOS, = As,S, + H,S + 2H,0; the disulphoxy-compound thus: 6H,AsO,S, = As,S, + As,S, +- 4S + 2H,AsO, + 6H,0. The three sulphoxyarsenic acids are not precipitated at once by Weinland’s! reagent, which fact serves to distinguish them from sulpharsenic acid. All three can be readily separated from arsenic acid by means of magnesia mixture which precipitates only the latter. The formation of these three sulphoxyarsenic acids, their instability and the products of their decomposition, account in a perfectly rational manner for all the irregularities accompanying the interaction of sulphuretted hydrogen and arsenic acid. This is a summary of our work so far as it has thus far progressed. Princeton, N. J., April 4, 1904: 1 An aqueous solution of tartar emetic and Rochelle salt. ~ 116 RICHARDS—THE ATOMIC WEIGHT OF NITROGEN. [April7, SOURCES OF ERROR IN THE DETERMINATION OF THE ATOMIC WEIGHT OF NITROGEN. (Contribution from the Chemical Laboratory of Harvard College.) BY THEODORE WILLIAM RICHARDS. (Read April 7, 1904.) The combining weight of nitrogen presents a problem of un- usual interest, because of the uncertainty which still clings to it, in spite of the careful work of some of the most accurate of chemical experimenters. Uncertainty of this kind implies a lack of com- prehension of some unknown variable or variables, and it is always possible that the determination of these variables may lead to the discovery of some new important fact or principle. Thus the accurate work of Lord Rayleigh m demonstrating that the less active gases of the atmosphere are somewhat heavier than pure nitrogen, led to the discovery of argon and the other inert gases. The data for computing the atomic weight of nitrogen are mani- fold, because nitrogen enters into many well-defined compounds. Unfortunately, however, it is always necessary to find the weight of the nitrogen indirectly. The most extended series of experiments was instituted by the great Belgian chemist Stas, who attacked the problem in various ways, converting silver into the nitrate, con- verting this nitrate into chloride, converting the nitrates of potas- sium and sodium into chlorides, and comparing ammonic chloride and bromide with pure silver. The average results of these experi- ments have been variously computed, the extreme estimates of the atomic weight of nitrogen ranging between 14 039 and 14.058 if the atomic oxygen is taken as 16,000.! ' The early work of Stas involving argentic chloride must all be rejected, because insufficient precautions were taken concerning its solubility. Among the other pertinent data obtained by him the following, easily traced in Clarke’s convenient Recalculation of the Atomic Weights (1897), seem to me the most important. (10 == 16,000; A, = 107,930; Cl = 35.455, H = 1.0076, Br = 79.955) 100,000 parts of silver gave 157.478 of its nitrate....., piediviy abe N == 14.036 100,000 parts of silver correspond to 49.599 of ammonic chloride, .N == 14,047 100.000 parts of silver correspond to 90,830 of ammonic bromide, ,.N = 14.048 Difference between molecular weights of alkaline nitrates and chlorides = 26.589.......... o:6'e'b po ha be oleiee oelkceaiaaa och és N = 14.043 Marignac's work on argentic chloride and nitrate leads to a much lower value t 1904.] RICHARDS—THE ATOMIC WEIGHT OF NITROGEN. 117 Stas himself concluded from these results that nitrogen was almost certainly higher than 14.03, and probably about 14.045, basing his conclusion upon a somewhat doubtful application of the theory of least squares.’ In spite of the great care taken by Stas in this unusually extended investigation, it is of course not impossible that small constant errors might have existed in parts of the work. Stas was by no means infallible; his long oversight of the solubility of argentic chloride, the uncertainty concerning the amount of oxygen occluded by his silver, and his frequent use of glass vessels somewhat attacked by his reagents for long-continued operations, being among the evidences that he too was mortal. Nevertheless, it is true that Stas was more precise than any one who preceded him; and his results cannot be overthrown without much conclusive experimental evidence. Three years ago the accuracy of one of these series’ of experi- ments made by Stas was impugned by Alexander Scott, namely, the series in which ammonic bromide was compared with silver. The atomic weight is computed from the result of these experi- ments by subtracting the weight of the bromine precipitated as silver bromide from the weight of the ammonium bromide, in order to find the weight of ammonium present. Because the bromine is equivalent to the silver, the ammonium previously united to the bromine must also be equivalent, and upon assuming the atomic weight of silver to be 107.93, the molecular weight of ammonium is easily found to be 18.078. Subtracting from this four times the atomic weight of hydrogen, that of nitrogen remains—namely, 14.048— because ammonium consists solely of nitrogen and hydrogen. Now Scott contended, with plausibility, that this particular spe- cimen of ammonium contained impurities, substances other than nitrogen and hydrogen, because Stas admits that his bromide was not perfectly colorless. These impurities would all be included in the estimate of the for nitrogen (14.01 to 14.02), a discrepancy which it is not easy to explain, unless the chloride was precipitated from a solution so concentrated as to occlude nitrate, The lack of details in his description makes it impossible to decide this question. 1 Stas, Untersuchungen (Aronstein), p. 322 (1867). 2 J. Chem, Soc, Trans.,79, 147 (1901). 118 RICHARDS—THE ATOMIC WEIGHT OF NITROGEN. [April7, weight of the nitrogen, since this is merely the remainder left after subtracting that of the bromine and the hydrogen. Hence, Stas’s estimate of the atomic weight of nitrogen is probably too high— how much too high it is impossible at once to decide. Some of the possible impurities have so powerful a color that an inappreciable weight of them might darken visibly an otherwise pure sample of salt; but an inappreciable weight could not affect the combining proportion, hence the error may be negligible. In any case it is obviously well that Stas’s experiments on ammonic bromide should be verified ; and the repetition was under- taken by Scott. His ammonium preparation was purified, as he states, by drastic methods, and was beautifully clear and colorless. There is little doubt that, so far as the ammonium was concerned, the salt was purer than Stas’s. Unfortunately, however, in his anxiety to purify the ammonium, Scott evidently neglected to purify adequately the bromine which he combined with it. His own results prove this fact indubitably, for he found on the aver- age’ that 107.93 parts of silver combined with only 79.943 parts of his bromine,’ a figure perceptibly lower than the most probable value, 79.955. ‘This latter value, computed from Stas’s work, has been repeatedly verified in the Chemical Laboratory of Harvard College during the last twelve years with only very slight variations. In order to show how definite the figure is, there is given belowa table of all our most refined recent work on this ratio.* 1 Neglecting one imperfect experiment, 7. Chem. Soc. Trans., 79, 147 (1901). ? Even this weight of bromine may have been too high, since Scott apparently overlooked the danger of the inclusion of water by the argentic bromide (Proc. Am. Phil. Soc. 1903, 28). The fused substance is the only safe standard of reference, In this connection it should be noted that an impurity in Scott’s silver would have caused an error in the same direction, Scott seems to have taken more pains with his silver than with his bromine, although indeed he con- demns on the basis of a single analysis a method of fusing it which many others have found satisfactory (/. Chem. Soc. Trans., 79, 15, 1901). ® Clarke’s similar table, Reca/c., p. 46 (1897), is necessarily incomplete and includes some imperfect preliminary results, besides containing several minor mistakes in calculation, 1904.) RICHARDS-—-THE ATOMIC WEIGHT OF NITROGEN. 119 ini P i- eference Proc.) Per cent. of sil- Spree caren {1 Ne: So Sep eth Aaktyet.: PRE ae "ver in Apter. RD as 52 cis's 3 weceee-| Last seven | Richards 28, 28, 29 57-444 SN hers eS Seven “ 30, 389 57-444 I i) mathe, mactina'gre Last series “ $1, 178 57-445 Cl BAA See erse Last seven | Cushman 33, III 57-444 COM A ai5 sececee| Last five Baxter 33, 127 57-440 RE ns cdianent dive ua Three Merigold 87, 393 57-447 REE ica esis eee ated Two Richards 28, 17 57-445 REE. ee APPS as One “ 80, 380 57-440 Te avinceeeeee “ 81, 165 57-444 ERE RRA TR: One Cushman 33, 106 57-445 ea See eae One Baxter 33, 122 57-444 eR LEE wladiesiel wo Baxter 34s 353 57-447 FVGTORC, . ocbevilaccnec SRheWetaed aee tink sos eeee 38 3.3629 0.6394 183.41 ‘« These numbers, in that they indicate the atomic weight of tung- sten, are worthless; in that they show promise for the new method, » are of value. The presence of impurity would lower the result ; what value the method will give for pure material can only be con- jectured.’’ Taylor’s experience re-emphasized the absolute necessity of satis- fying ourselves beyond every reasonable doubt that the material for the atomic weight determinations was pure ; at least as pure as the means at hand would furnish. The admission of Schneider that his purest substance contained traces of impurity, insoluble in caustic potash, and Taylor’s discovery that every sample of tung- sten trioxide tested by him gave a residue, insoluble in sodium carbonate, made us very solicitous regarding the purity of all material which had been used in any previous investigations, for it will be found upon consulting the literature that almost every experimenter was cortent to proceed with ammonium paratungstate which was perfectly white in color. Three to five recrystalliza- tions were held to be sufficient to attain that condition. Our doubts became so overwhelming that it was decided to begin the work anew with the mineral wolframite and to ascertain, once for all, what it contained in order that search might be made for all such constituents, and every effort put forth to insure their per- fect removal from the salts which might be experimented upon. Accordingly, in the summer of 1go1 large quantities of wolframite, from Lawrence County, S. D., were decomposed and the resulting tungstic acid converted into ammonium paratungstate. The mother liquors from the salt were black in color and gave in due time the interesting compound—ammonium vanadico-phosphotungstate— described by us in the Jour. American Chemical Society, 24, 573. Its discovery added, of course, vanadium and phosphorus to the list of possible contaminating substances: columbic oxide, silica, 4 132 SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, molybdic oxide, ferric oxide, manganese oxide, etc. Large amounts of ammonium paratungstate were taken for the purification of the tungsten trioxide by methods.which may now be presented. Method r.—This first fraction was collected apart, dissolved in distilled water and recrystallized ; the first crystallization was again set aside and dissolved, this operation being repeated ten times. A portion of the sixth crystallization was ignited in a platinum crucible and the resulting oxide was digested on a water-bath with a 2%, solution of sodium carbonate free from iron. The oxide dis- solved, but its solution was quite turbid. Upon standing, a white residue settled out, which after washing with water and decomposi- tion with a few drops of hydrochloric acid showed tungstic acid © and iron. Portions of the tenth recrystallization behaved similarly. The mother liquors, including that from the tenth crystallization, assumed a dark-brown coloration upon concentration, indicating thereby that not only iron, but that also the vanadico-phospho- tungstate, already alluded to, continued with the ammonium tung- state. Vanadium too was found in a portion of the tenth crystal- lization when it was heated in an atmosphere of hydrochloric acid gas. Hence it was concluded that this method was unsatisfactory and it was abandoned. Chemists, who in the past were content to look upon ammonium paratungstate as being very pure when its color was perfectly white (and from published statements most have been content with this criterion), which it is after the third or fourth crystallization, cannot have had pure substance for their investigations, hence the fluctuation in their results is easily com- prehended. Method 2.—This may be called the method of Borch. In it the mineral was fused with sodium carbonate, the fusion exhausted with water and after filtration the liquid was precipitated with calcium chloride. ‘The resulting calcium tungstate was filtered, washed and decomposed with hydrochloric acid. The liberated tungstic acid was dissolved in ammonia water and the ammonium salt crystallized out. The method was carried out with rigid adherence to the printed instructions, but it proved entirely unsatis- factory. Method 3.—Solutions of ammonium paratungstate, decidedly brown in color, were boiled with precipitated calcium carbonate. Ammonia and carbon dioxide were evolved, while calcium tung- state was precipitated, The solutions lost nothing in color. After 2 a 1904.] SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 133 filtering ont and washing the calcium tungstate it was boiled with hydrochloric acid. Tungstic acid of a rich yellow color separated. It was washed and dissolved in ammonia water. ‘There was a very slight bluish-colored residue. ‘The ammonium paratungstate, which crystallized out, was perfectly white in color, but a portion of it ignited and the resulting oxide, digested with a dilute sodium car- bonate.solution, disclosed the usual white residue in which tungsten and iron were found. The salt, purified in this way, was fully as free from impurities as a salt which had passed through six crystallizations from water. This fact led us to prepare pure calcium carbonate and hydrochloric acid before repeating the method with another portion of ammo- nium paratungstate. Commercial’ calcium carbonate was boiled with a solution of pure ammonium chloride. Iron and other impurities remained with the excess of calcium carbonate. The filtrate from the latter was precipitated with pure ammonium carbonate. The resulting calcium carbonate was thoroughly washed and then dried. __ Ordinary hydrochloric acid was saturated with calcium chloride and after the addition of phosphorus pentoxide was distilled with water. This gave pure acid. With these purified reagents ammo- nium paratungstate, which had passed through several crystalliza- tions, was subjected to the treatment outlined in the beginning of this section. The purified salt, when tested, showed but.traces of impurities and it is very probable that these, after several repeti- tions, would disappear entirely. The actual trial was not made, because another course seemed to lead to the desired end in a much shorter period. Experience also showed that nitric acid was preferable. Method g.—In this procedure a boiling solution of ammonium paratungstate was decomposed with hydrochloric acid. The pre- cipitated tungstic acid was again dissolved in ammonia and the decomposition repeated. Repeating this process several times yielded tungstic acid which might be asserted to be quite pure, although when the mother liquors from the several fractions of the ammonium salts were reduced to a small volume the dark color appeared. A white residue, although slight, was also obtained from the ignited tungsten trioxide. Wyman (Thesis, University of Pennsylvania, 1902) found, after twenty-five decompositions with hydrochloric acid, evidences of similar contamination. This 134 SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, chemist tried seven different schemes in his endeavor to obtain pure tungstic acid without the desired result. Accordingly, on resuming this part of our study in the summer of 1902 we deter- mined to eliminate every possible source of contamination from the various reagents which we proposed using. Thus, fifteen liters of hydrochloric acid were purified as already described. It was free from every impurity, which was proved by carefully repeated tests. Eighteen liters of nitric acid were distilled after the addi- tion of pieces of pumice and some sodium hydrogen phosphate. The product left no ponderable residue when a definite volume of it was evaporated to dryness in a platinum vessel. The sodium carbonate was made pure by fusing it in a platinum vessel and introducing into the molten mass a small quantity of pure, precipi- tated calcium carbonate which dissolved ; the mass being held for five minutes in the liquid state. After cooling, the fusion was allowed to dissolve out in cold water. The calcium carbonate, iron, etc., were filtered off and the solution evaporated to crystal- lization. The sodium carbonate which separated was recrystallized four times. Platinum vessels were used for the purpose. They are essential. Portions of the purified salt were examined for silica and iron, and their absence demonstrated. There was now every reason to believe that the reagents, including the water (for it had been redistilled), were pure. They contained nothing which would contaminate the tungsten trioxide. Therefore, if the latter left a residue upon digestion with a dilute sodium carbonate solution, that residue plainly came out from the tungstic oxide, And now we must digress a little. Wyman experienced difficulty and annoyance in his efforts to dissolve tungstic acid in ammonia water, Others have had, doubtless, similar experiences. There invariably remains a bluish-white mass which no amount of ammo- nia or protracted boiling eliminates. More than a kilogram of this substance had accumulated from Wyman’s work and came into our possession. Its bluish-green tint suggested the presence in it of ‘some reduction product, probably occasioned by the hydrochloric acid. We accordingly projected the mass into boiling concen- trated nitric acid. A violent evolution of chlorine immediately followed and continued until the material had acquired a rich yellow color. Whence came the chlorine? More of the blue residue was dried at 100°, then heated upon a platinum foil. Great volumes of ammonium chloride were expelled, leaving pale-yellow f x st ee Oe oe 1904.) SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 135 colored tungsten trioxide. An analysis of a portion of the dry, bluish material was made, when 5.48% of NH, and 9.65% of Cl were found. The presence of this amount of foreign substance could only be accounted for on the assumption that in the libera- tion of tungstic acid from ammoniacal solutions of ammonium tung- state with hydrochloric acid portions of the latter and of the ammonium chloride were so combined that no-amount of washing would remove them. They are retained, and firmly, by the tung- stic acid. et those who question this examine the white, slimy residue which appears on attempting the solution of tungstic acid in ammonia. Most of us have quieted our consciences on this point by asserting that such residues are ‘‘ those persistently insolu- ble paratungstates.’’ If the solution from such residues be repre- cipitated with hydrochloric acid quantities of insoluble bodies again appear. These are beyond question ammonium chlorinated tung- stic acid derivatives which even prolonged boiling with concentrated acids will not change to the yellow acid. They merit further study. The preceding experience emphasized the necessity of removing all the ammonium chloride in tungstic acid if the precipitation method of purification was to be pursued. This was done in the following way: two to three liters of concentrated nitric acid, diluted with water to half their volume, were heated in a large porce- lain dish until it began to fume strongly, when 25cc. of pure hydro- chloric acid and three kilograms of dry and fairly pure ammonium tungstate were added. Vigorous action set in and volumes of decomposition gases escaped. The mixture was constantly stirred during the operation. As soon as the action diminished, ten to fif- teen cubic centimeters of pure hydrochloric acid were introduced at intervals. As the decomposition approached completion, the yellow tungstic acid lost its porous character and collected as a heavy granu- lar powder upon the bottom of the dish, but it was heated with occa- sional stirring for a period of from three to four hours, At the expiration of that time, there remained only tungstic acid and nitric acid with traces of chlorine and ammonia. The tungstic acid was washed by decantation with pure distilled water until the tung- stic acid suspended in the solution subsided very slowly, and the wash-water from it showed but a faint acid reaction. The subsided yellow-colored acid was placed in a porcelain dish, hot distilled water was poured over it and ammonia was conducted into the solu- 136 SMITH AND EXNER—ATOMIC WEIGHT OF TUN @STEN . [April 7, tion. As a rule all of the acid dissolved, and there was at the most a very small residue. Thus forty-five liters of a saturated solution contained a residue which weighed less than two grams. The ammonium paratungstate separating from such a solution showed the needles and plates characteristic of that salt. Only the first three fractions were preserved. They represented seven-. eighths of the entire substance. The other portions were set aside. A second and a third treatment, as outlined above, was given the first three fractions, and when the salt, finally obtained, was sub- jected to the sodium carbonate test, allowing the solution to stand over night, it remained absolutely clear. The mother liquor from the salt, reduced to five cubic centimeters, remained colorless. One object in this long and baffling study had at last been obtained. We were the possessors of seven hundred grams of pure ammonium paratungstate. Three kilograms of impure ammonium paratungstate were decom- posed by acids as described above, the operation being repeated five times, when the ammonium salt from the last decomposition responded admirably to the crucial tests. This salt disclosed none of the substances which were found accompanying the tungstic acid originally, hence the latter was considered pure and was applied as will appear in subsequent paragraphs. However, before proceeding further it seems proper to direct attention to certain other experi- ences which possess interest and value. Tue RESIDUE OBTAINED BY DIGESTING TUNGSTEN TRIOXIDE WITH SopIuM CARBONATE. Ammonium paratungstate, after three crystallizations and when quite white in color, was ignited in a platinum crucible. Two hun- dred grams of oxide were obtained in this way and were digested with a 2% solution of sodium carbonate of excellent commercial quality. Quite a residue appeared. It was washed and dried. A portion, weighing 0.4712 gram, was digested with agua regia. The insoluble part weighed 0.4309 gram, while the solution of the soluble constituents, weighing 0.0403 gram, was reduced to a small volume, diluted with water and precipitated with ammonia. The iron weighed 0.0337 gram as ferric oxide. Manganese and plati- num were also found, The tungsten trioxide, when digested with hydrofluoric acid, lost 0.0149 gram, representing silica. Or, if the results be tabulated, they show: 1904.] SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 137 Weight of residue ..65...5...: Ate | es re Mh LN EOS RT ee Cees 0.4160 « 88.28% eal wikis lnn.o.ae one alecwhs's 0.0149 * 3.16 « Fae CMe ain Sais m.ee-¥% aie 0 08 acos Gugay * 7-15 Manganese, platinum, etc. ............ has) Si leg Mle aa mere Apap aie ‘ 98.59% Even pure sodium carbonate will not remove all of the impuri- ties, although it may serve to test the purity of the oxide as to the iron, etc., which may be present. IGNITION OF AMMONIUM PARATUNGSTATE. The ignition of this salt in platinum vessels, as ordinarily con- ducted, contaminates the trioxide with platinum. To minimize this contamination a platinum crucible was fitted tightly two- thirds of its length into an asbestos board. A platinum wire shaped into a tripod was set upon the bottom of the crucible. A smaller platinum crucible was supported by the tripod. Into the latter were introduced from time to time not more than from two to three grams of ammonium tungstate. A red heat was applied to the outer crucible. The ammonia was expelled in the course of half an hour, when the crucibles were covered with an inverted porcelain lid, it being lifted from time to time to admit air. Constant weight was obtained in two hours. This procedure gave the best results which could be gotten by the use of platinum crucibles. While the oxide is cooling it should be protected from all reducing atmos- pheric dust, because the hot oxide is extremely sensitive to the action of such substances. This is evident from the following: a platinum rod previously heated in a flame and applied to the hot oxide produces no change, but if the rod be touched quickly to the skin and then laid on the hot oxide, a green spot will appear at the point of contact. The efforts to substitute silver and gold crucibles for those of platinum demonstrated that these metals, too, were appreciably absorbed by the oxide. Porcelain crucibles were used, notwith- standing the absorption of silica, which would of course become greater as the time of ignition was prolonged and as the heat was increased. Further, the oxide in immediate contact with the porcelain invariably showed a green color. The glaze of the crucible always indicated etching. With an unglazed crucible the action was not so evident, hence the contamination was not so great, and the most satisfactory results were gotten by setting the 138 sMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN, [April 7, / unglazed porcelain crucible in a platinum crucible and bringing about the ignition of the salt in this double-walled chamber. The coloration of the surface of the oxide was extremely slight. Experi- ence, however, eventually showed that the best course to pursue con- sisted in digesting the ammonium paratungstate directly in a porcelain casserole with pure nitric acid and a few cubic centi- meters of hydrochloric acid until it was completely decomposed, and the ammonia and hydrochloric acid were destroyed. When the tungstic acid was evaporated to complete dryness, it showed a rich orange-yellow color. It was transferred to an unglazed porcelain crucible and there ignited gently for half an hour. This may be done over a direct flame, the crucible being covered with an inverted porcelain lid. Any enclosed nitric acid was expelled by. the gentle heat, and the weight soon became constant. The result- ing tungstic oxide had a uniform yellow color. Green was abso- lutely absent. This procedure eliminated the reduction caused by the ammonia, and it may be added that by its use glazed crucibles were employed every day in similar ignitions for several weeks with- out showing the least etching or corrosion of the surface. Having at last gotten pure salt and pure oxide, the question arose as to what method should be adopted in the determination of the atomic weight of the metal. The method proposed by Taylor (p. 130) was new. ‘The results he obtained were with material not especially purified, yet their fair agreement pointed to the possibil- ity of arriving at a definite value with the pure substance, such as was now available. Preliminary trials were executed according to Taylor’s suggestions, using glass apparatus just as he had done, and drying at 400° to constant weight. The weighings were all made on the same day,.and under uniform conditions. The main pur- pose was to ascertain whether concordance in results could be realized. The results in the subjoined table show the opposite: Na,CO, WO, CO, At, Weight Bee dsr 9 b> «+ 5-9 gm. 2.45645 gm. .46775 gm. 183.07 Wetasesishict 56 2.72292 51785 “ 183.36 EATER EE 55 “ 332953“ 63288 « 183.48 Bits Beet ei ee 4-7 #8 3.96720 * 75473 183.29 Rivage aes ber 40 * 3.44944 “ 65489 “ 183.75 Ginsicweonevaus 41 4 3-41273 “ -647960 183.74 Tevevcanevsers 438 6.10309 1.16087 183.32 Megs scers es) s 3.9 4% 6.39735 “ 1.21644 “ 183.39 Qincccsvesepes 35 “ 2.17450 “4 41332 183.48 BD's ci ond e'us ee 31 4 1.57903 * .29966 « 183.85 ~ © 1904.) SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 189 The trioxide used in experiments 1 to 4, inclusive, was obtained by gently igniting the ammonium salt in a porcelain crucible. That used in experiments 5 and 6 was strongly heated in a porce- lain crucible. In -7 and 8 the ammonium tungstate was heated for one hour in a double platinum crucible. The oxide in experiment 7 had been heated two hours in the same kind of crucible, while in experiment 10 the ignition continued for two hours in the double platinum crucible. The gradual rise in the value to 183.85 by protracted heating could surely not be due to the expulsion of vola- tile matter, for there was no change in weight after the first hour of ignition in the double crucible. Evidently the oxide absorbed impurities which led to the rise in the atomic weight. Accordingly, samples of ammonium paratungstate were ignited under conditions as nearly similar as possible in crucibles of platinum and of porce- lain. The values from the oxide in the crucibles of porcelain were higher than those from the oxide made in platinum crucibles, show- ing in all probability, that the oxide took up more foreign material from the porcelain than from the platinum. ‘Therefore, the mere ignition of the ammonium salt in vessels such as have been described drew in sources of error. These would, of course, have to be eliminated if the method was to be tested upon its own merits. It was sought to accomplish this by igniting thoroughly dry ammonium paratungstate and ascertain the loss (water and ammonia) sustained by different amounts, which resulted in dis- covering that the percentage of volatile matter could be obtained to within less than 0.01%, which would answer for the purpose of atomic weight determination. And therefore, in the actual experi- ments, the ammonium paratungstate was weighed out directly into the flask, it being only necessary to make the proper calculations to arrive at the amount of trioxide which was thus used. Six deter- minations were made; the results in the atomic value varied trom 183.4 to 183.81. The early explanation for the lack of concor- dance, if the method was not faulty, would be to suppose that the action of the soda upon the glass would withdraw varying amounts of silica, and there would follow, of course, the liberation of corres- ponding amounts of carbon dioxide. If this really occasioned the error, it was hoped that the substitution of a platinum bulb, similar to the glass vessel, would lead to success. This was done. The experiments were performed as before, with slight modifications where it was considered advisable and advantageous. The atoinic 140 sMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, values in a series of five trials ranged from 182.85 to 183.64. Patient search was made for the reason, every step being tested repeatedly, until eventually the conclusion was forced upon us that carbon dioxide, in varying amounts, was disengaged through the decomposition of the sodium carbonate in the final drying. Jac- quelain (Jahresb., 1860, p. 116; A. Ch. [4] 28, 86, and A. Ch. [3] 32, 205) showed that this salt loses from 0.03 to 0.05% in weight at 400°, and other observers have shown that the loss con- tinued with the length of the period of ignition and with the temperature. Here, then, was a serious defect in the method which would explain why the values found were low, and why they differed so widely. The attempts to correct this weak point proved futile, so that the method, having had a thorough trying-out, was abandoned after months of arduous work. . It was hoped that perhaps a normal silver tungstate might be made, which after solution in potassium cyanide could be electro- yzed and the value of tungsten obtained from a comparison with the precipitated silver. Fifteen experiments were made. In one series (the best) of five experiments the results varied from 184.00 to 184.39. It was found, after much search, that there could be no certainty as to when a normal salt was really in hand. Washing and drying, even when performed with the utmost care, occasioned a change in the character of the salt. The method was discarded. An effort was also made to obtain a cadmium salt of definite composition. Much time was given to it, and experiments were made in the electrolysis of bodies believed to be uniform in com- position. The atomic values ranged from 181.90 to 185.71. Having subjected three new methods to vigorous tests in our efforts to solve the problem along new lines, and having found them utterly deficient, the hope still remained that possibly some of the earlier methods might with pure material give satisfactory results. The writers felt, without meaning to reflect in the slightest upon earlier investigators, that their material possessed the merit of superior purity; and if that were really the case, older methods, simple in principle and easy of execution, might well be expected to give concordant values. Of the 175 experiments made by the entire corps of previous investigators, there is but one short series, namely, that of Pennington and Smith, in which there is that degree of concordance which is desirable and necessary in fixing the 1901.] SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN, 141 atomic value of any element. Splendid as is the work of Schneider, worthy as it is of high praise, there still remains the fact, not to be pushed aside, that between the minimum and maxi- mum values there is a difference of more than a unit. The atomic value given by Schneider, Hardin and others for tungsten is 184—the mean of very widely differing series. Cognizant of these facts, with faith in the greater purity of our material, steps were taken to repeat several older procedures. PREPARATION OF TUNGSTEN HEXACHLORIDE. Chlorine, free from exygen and moisture, is absolutely essential to obtain this compound pure and in comparatively large amounts. The product must also be sublimed repeatedly in an atmosphere of chlorine, without exposure to the air. The first condition, although apparently simple, is really very difficult to attain ; and after much experimenting, we cannot say that we got chlorine absolutely free from moisture. But the quantity of oxychloride formed along with the hexachloride may be taken as an index of the amount of moisture (also oxygen) in the chlorine. The generator was charged with material sufficient to yield chlorine an entire day without the addition of acid and consequent introduction of air into the apparatus. When the flow of gas com- menced to grow less only the gentlest heat was applied for a few minutes to the generator. The chlorine was most completely dried by conducting it through three six-inch U-tubes connected in series, containing pumice stone saturated with pure concentrated sulphuric acid, and in the bend sufficient acid to fill the bottom of the tubes, thus causing the gas to bubble through the acid before each new preparation. Indeed, it was about every fourth day that a renewal was made. Only traces of oxychloride were observed. The reaction of chlorine and metal took place in a combustion tube of soft glass, 15 to 18 mm. in diameter and 4% feet in length. The tube was contracted in two places to the thickness of a lead pencil, thus making three sections, of which the first was 3 feet in length, the second 1 foot and the last % foot. A porcelain boat carried the tungsten metal. The chlorine was passed through the apparatus for two hours before any heat was applied. This was done to expel the air. Then the burners of the furnace (to within three of the boat) were lighted, beginning with those most distant from the boat, the flames being small. The tube beyond the boat 142 sMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, thus reached a temperature of nearly 350° C. These burners were then extinguished, while those immediately beneath the boat were lighted, the flames being small. The tube to a length of eight inches beyond the boat was also heated. In a very short time the reaction began, noticeable at first in the yellow vapors which con- densed in the colder part of the tube, beyond the furnace, to the light brown oxychloride. This did not continue more than two minutes, when copious reddish-brown vapors appeared and con- densed beyond the lighted burners to brilliant blue-black needles. The formation is very rapid. The utmost vigilance is constantly required to the very conclusion of the experiment. In two hours twenty grams of metal may be fully converted into hexachloride. That portion of the combustion tube at which the hexachloride condenses should be kept just hot enough to cause the traces of oxychloride to pass beyond the hexachloride. This can be readily adjusted after a little experience. When working with large amounts the deposits of the hexachloride may obstruct the tube. In that event manipulate a lamp flame with the hand beneath the chloride until it is partially melted. This converts it into a com- pact solid, requiring less space. Melting and resublimation of the chloride removes every trace of oxychloride. Two sublimations are sufficient for the purpose. The tube can then be sealed off at the contracted points. Perfectly pure hexachloride has a beautiful, brilliant steel-blue color. It can be readily transferred to clean, dry weighing bottles and preserved in them. It has marked stability. There is no perceptible action on bringing the chloride into water at the ordinary temperature even after considerable time. On the application of heat the decomposition does not begin until the temperature of the water reaches 60°. The specific gravity of the hexachloride, taken in water at the ordinary tem- perature, equaled 3.518. After all the weighings were made the water employed for the purpose was tested with litmus; it gave the very faintest acid reaction. Having obtained in the above man- ner large quantities of the hexachloride, it was decided to change it to trioxide and thus arrive at the atomic weight of the metal. Roscoe had determined the chlorine in thiscompound. His results were not especially concordant. Perhaps this was due to the involved method or to the presence of traces of oxychloride: However, our thought was to adopt the simplest available course, hence we aimed to convert the chloride into oxide. Roscoe has 1904.) SMITH AND EXNBR—ATOMIC WEIGHT OF TUNGSTEN. 143 stated that when hexachloride is directly decomposed with water and the resulting acid ignited to oxide, the latter will contain chlo- rine which cannot be expelled by heat. We had hoped to pursue this method, but as it had the condemnation of so high an authority the hexachloride was introduced into freshly distilled ammonia water, contained in a weighed platinum dish, with the expectation of eventually getting ammonium tungstate and chloride which would leave the trioxide upon ignition. Experience showed that the quantity of the resulting ammonium chloride was so great that even with the most careful ignition there was much danger of expelling mechanically appreciable amounts of the oxide. Nor was it forgotten that it is very doubtful whether from such a mix- ture the chlorine could be completely removed by heat. The treatment of the hexachloride directly with nitric acid was also found impracticable. In spite of Roscoe’s objection to the decomposition with water it was believed that the transposition could be carried out. Five glazed No. 2 porcelain crucibles of 40cc. capacity were selected, thoroughly cleansed and ignited, allowed to cool in vacuum desiccators and weighed upon a specially constructed Troemner balance, sensitive to 5 of a milligram. There was next introduced into each one of them tungsten hexachloride from a weighing bottle which was reweighed after the removal of each portion. The crucibles with their chloride content were placed on water-baths and cold distilled water introduced into each. When the volume of water was insufficient for the quantity of chloride sufficient heat was generated by the reaction to make the water boil and spattering followed. At 60° the decomposition proceeded quietly to the hydrated trioxide, which at the beginning had a slight greenish-yellow color, due probably to imperfect decomposi- tion, as mentioned by Roscoe, but this tint disappeared as the hydrochloric acid was expelled. When the mass was perfectly dry a few drops of pure concentrated nitric acid was introduced from a pipette upon the trioxide. Instantly any green tint vanished and was replaced by a rich orange-yellow color. The excess of nitric acid was slowly evaporated away and the oxide assumed a pale yellow hue. The crucibles were now removed from the water-bath, and after careful drying were ignited for half an hour to a dull red heat, then allowed to cool in the desiccator, and at the expiration of an hour and a half were weighed. 144 SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, In the calculations the values for oxygen and,chlorine were taken at 16 and 35.45 respectively. The specific gravity of tungsten tri- oxide was found to be 7.157 and that of tungsten hexachloride 3.518. Seven different series of determinations were made, each from a different sample of hexachloride, The results appear in the sub- joined table: .of | No.of |Wght. of WCle/Wght. of WOs) at went. f f TBixp. | Series, | Cot. for'Vac. | Cor. for Vac.) “AS cP" | "Screg | Means I 3.18167 1.86085 184.04 2 a, 2.66612 1.55903 183.94 184.01 3 3-5 2632 2.06244 184.05 4 j 1.52117 0.88972 184.07 1.22299 0.71523 184.00 3 II. 2.28445 1.33603 184.01 184.04 7 3-25404 | 1.90337 184.10 8 3.37078 1.97133 184.01 9 III. 7.76488 4.54082 183.98 183.98 10 2.08764 1.22114 184.11 II 2.80141 1.63859 184.09 12 IV. 3.24328 1.89681 184.02 184.08 184.04 13 4.9797 2.91262 184.06 14 3.0403 1.77838 184.10 1 4.31046 2.52133 184.10 I 2.21201 1,29381 184.07 17 V. 2 70368 1.58135 184.06 184.06 18 3.60658 2.10934 184.03 19 2.63037 1.53835 184.02 20 3.41668 1.99808 184.07 21 3.49940 2.04675 184.06 22 VI. 3.86668 2 26145 184.05 184.04 23 3-40202 1.98970 184.03 24 3.20661 1.87533 184.01 35 3.26386 1.90909 184.09 : VIL 73 383 3-94031 183.94 184.06 27 7-37 4+31643 184.14 It should be mentioned here that at the conclusion of these experi- ments etching or corrosion of the glaze of the crucibles could not be observed. Nor was there any stain upon them; they looked as if they had been unused. i emi 1904.] SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 145 METALLIC TUNGSTEN. It would be superfluous to set forth here the steps taken in pro- curing the metal. They are familiar to every reader. They were identical with those described by Hardin. One point, however, is worthy of notice. It was discovered that if the trioxide, reduced to metal, had been previously gotten by the ignition of ammonium paratungstate in vessels of platinum, then it might well be expected that after the removal of the tungsten from the reduction boats the latter would show dark spots here and there. This occurred, but uncontaminated trioxide was repeatedly reduced in porcelain by hydrogen without leaving dark stains. Several experimenters—Riche, Desi, Shinn and Hardin— endeavored to reach the atomic value of tungsten by collecting the water resulting from the reduction of definite amounts of trioxide in hydrogen. Their results were disappointing in the extreme, although the method is surely rational and in some measure ideal. The reasons for its failure have never been satisfactorily explained. We were induced to give it trial. Every attention to detail was scrupulously observed. The results were most disappointing, and yet we cannot give a reasonable explanation for our failure. There seems fo be an inherent defect in the method which we were unable _ to lay bare. We also reduced definite quantities of trioxide to metal, and from the loss in weight sought to get the atomic value of tungsten. Again the results were discordant. The boats were never stained from the reduction, nor was the porcelain tube in which the reduction took place stained, but on close scrutiny particles of metal could be seen along the sides of the tube. They rested loosely upon it and were removed with ease. This metal, in all probability, was carried out into the tube by the aqueous vapor produced in the reduction. This is, therefore, a serious point in this method. There remained, finally, only method 2, another time-honored method, upon which much discredit had been cast. Yet with pure material it seemed worth the while to give it further trial. The metal used in this study was made from trioxide obtained from the hexachloride. Portions of it were weighed out into the same crucibles which had been used in the experiments with the hexa- chloride and gently heated with air contact. The steps in the PROC. AMER. PHILOS. SOC. XLIII. 176. J. PRINTED JUNE 4, 1904. 146 sMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. [April 7, ignition were those which any careful analyst would observe, so that they need not be mentioned here. The final oxide was uni- formly yellow in color throughout its entire mass. The weighings here, as in all previous experiments, were reduced to vacuum standard. The value of oxygen was placed at 16. The specific gravity of the oxide was, as before, 7.157, and that of the metal —19. In the appended table it is to be understood that each single series was made from portions of the same sample of metal. The results are: No. of No.of Weht. of W. |Wght. of WO,;) At. i Means of Mean of Exp. Series. in gms. in gms. of W. Series. Means. I & 2.24552 2.83144 183.96 183.96 3 2 Il. 1.78151 2.24619 184.07 184.07 3 1,63590 2.06270 183.98 4 III. 1.38534 1.74665 184.04 184.04 5 1.29903 1.63774 184.09 6 2.01302 2.53781 184.12 7 2.18607 2.75632 184.01 8 TV. 2.36755 | 2.98478 184.12 184.09 9 1.94958 2.45781 184.12 10 4.43502 5.59141 184.09 11 2.37603 2.99548 184.11 12 Vy 2.58780 3.26260 184.08 184.10 184.065 13 2.58503 3.25886 184.14 14 2.38298 3.00441 184.06 15 2.05578 2.59169 184.1 16 VI. 760828 4.34915 184.08 184.11 17 6.22621 7.84949 184.11 18 VII. 5.28444 6.66239 184.08 184.08 19 VIII. | 399095 5.03138 184.12 184.12 20 | IX | 7.30166 9.20647 184.00 184.00 ai ) | 3.44143 4.33870 184.10 aa | «=X. «=| :~«2.67709 | 3.97542 184.01 184,08 23 | | 4.96735 6.26229 184.13 In series VII, a very large quantity of oxide was heated in hydrogen from 9 A.M. until 5 P.M. The resulting metal was placed over night in a desiccator, and on the following day a por- LEE Rr Pe ra ea 1904] SMITH AND EXNER—ATOMIC WEIGHT OF TUNGSTEN. 147 tion of it was weighed out for the eighteenth experiment, the remainder being heated for a day more in hydrogen. After stand- ing over night a second portion was removed and used in experi- ment 19. The remainder was exposed all of the third day to the action of hydrogen, and was then oxidized for experiment 20. Had not the first reduction been complete, the results would not have agreed so well. The mean atomic value from the hexachloride is 184.04, that from the oxidation of metal 184.065, or the average of the two independent series is 784.05, which probably approximates the truth very closely and may be safely regarded as the atomic weight of tungsten. SUMMARY. Our study, extended over so long a period, has revealed— 1. That it is quite doubtful whether any chemists who in the past © occupied themselves with a determination of the atomic weight of tungsten have worked with pure substance. Tungstic acid is prone to form ‘‘complexes.’’ It was found that if the acid contain no iron, for instance, but be digested with acids, ey., hydro- chloric or nitric acid, in which iron is present, the latter will enter the tungstic acid. Iron and manganese are eliminated from the acid with the greatest difficulty. In the earlier work there is no evidence of their removal. Neither do we discover that vanadium and phosphorus had been considered as present, yet in purify- ing ammonium paratungstate by recrystallization alone it was found that the tenth recrystallization showed vanadium. 2. The slimy, greenish or bluish-white masses believed to be ‘¢ paratungstates ’’ because of their great insolubility are probably “complexes.” 3. The fourth method of purification can be relied upon to yield pure tungstic acid. 4. The use of pure sodium carbonate (2%) to dissolve tungsten trioxide gives an excellent means of ascertaining when the iron, manganese and silica are fully removed, but that its development into a method for the determination of the atomic weight of tung- sten is not at all probable. 5. The plan of digesting pure ammonium paratungstate with nitric acid, then evaporating to complete dryness and gently ignit- ing affords pure oxide. 148 MASON—RIPENING OF THOUGHTS IN COMMON. [April9, 6. That porcelain vessels are preferable to those of gold, silver or platinum for the ignition of ammonium paratungstate and tung- stic acid. = 7. That the oxidation of metal (method 2) leads to reliable atomic numbers when the material is pure. 8. That tungsten hexachloride can be completely transposed into pure oxide with water and a little nitric acid. The John Harrison Laboratory of Chemistry, University of Penna. THE RIPENING OF THOUGHTS IN COMMON. “Common Sense is Thoughts in Common.” BY OTIS T. MASON. (Read April 9, 1904.) Those who are entangled in official or commercial life, and, indeed, observant persons generally, will recall many instances in their daily experiences when they have mentioned a name only to see its owner appear. Or they have a friend, say, in the Straits Settlements. After a long silence they begin to worry about him and sit down to write to him. While they are thus engaged the postman hands in an epistle from Singapore signed with his name. There is, of course, an element of chance in such coincidences. A vast number spring out of deep-seated, normal biological condi- tions. It is not here denied that many, associated with abnormal or hypersensitive conditions, are so startling in time and detail as to give rise to beliefs in telepathy. Leaving out the causes just mentioned, this paper will be confined to those artificialities of life called culture, though the natural causes mix freely with these. The purposeful actions of all humanity have become so artificial- ized as to make the natural, physical man subservient to the new man, the lomo sapiens. Racial activities and community experi- ences have entirely changed, so that coincidences in speech, man- ners, customs, and arts, however surprising they may be, are also due to the maturing of thoughts, desires and purposes held in 1944.) MASON—RIPENING OF THOUGHTS IN COMMON. 149 common. And such agreements are not exceptions but are num- berless. Similarities and simultaneities in actions and thoughts among millions of persons form an unconscious never-ending drill, the activities passing imperceptibly from voluntarism into automatism. The coincidences of which notice is taken are not a drop in the bucket to the whole number. I shall speak of thoughts in common and the activities linked with them under the heads of dzology, speech, industries, fine art. social life, learning and lore, and religion. BIoLocy. To begin with activities that are purely biological, thoughts in common are shared with the animals. The revolution of the earth on its axis, producing day and night, causes nature to awaken in concert in the morning and to fall asleep in unison in the evening. There is no leader to the orchestra in the former, nor authoritative command or lullaby in the latter. With the change of seasons concerted movements of large masses of insects, fishes, birds, and mammals take place, lasting many days and extending over vast distances and spaces. Under other influ- ences hidden from our knowledge, the whole animate creation seems possessed of individual will only to work in obedience to a common will. This fact was observed three thousand years ago, for one of the Hebrew proverbs reads, ‘‘ The locusts have no king, yet they go forth all of them by bands ’’ (Prov. xxx. 27). This moving in concert has a more complex kind of action still, a sort of international code, existing between creatures of different species, genera, orders, families and even between the kingdoms of nature. It resembles a purposeful selection and is the natural fore- runner of altruism in culture. It is the hotbed of suggestion for the whole series of psychical mysteries. These maturings of thoughts in common are deep-seated in the human frame. ‘‘ As quick as a wink’’ does not mean a sudden period of capricious length; but one of duration as regular as the ticking of a watch. The physiologists, with their delicate appara- tus, have made wonderful discoveries in this direction. ll sorts of clever tricks are played on crowds successfully in the domain of psychology through these uniformities of action in biology. 150 MASON—RIPENING OF THOUGHTS IN COMMON. _ [April9, SPEECH. The first and easily overlooked occasion of thoughts in common is speech. Each word or phrase, and even whole sentences, have generic as well as specific meanings. Through the former they have acquired the habit of making like impressions on a multitude of minds or of calling forth identical responses. Through their specific, esoteric meanings they appeal to a smaller following, but more intensively. Every association or tribe has such formule, and their instanta- neous power of allaying the individual thought and merging the single into the organized opinion is a matter of common knowl; edge. Amid the multifarious capabilities of the vocal apparatus a small number of products are chosen, not by a committee, through laborious and purposeful efforts, but by the committee of the whole, which never adjourns. In some families of tribes, only the easy, musical, phonic ele- ments are picked out; while in others not far away, the harsh or guttural sounds are preferred. It has often been declared that these subtle combinations of breath- ings are more persistent than walls of brick and stone. This is not difficult to believe, since the verbal expressions that survive among a people body forth the imperishable thoughts and prejudices that long ago passed from the evanescent stage in the single mind to the fixed stage in the tribal mind. ‘The charge of plagiarism is fre- quently made by literary critics when the authors were totally unknown to each other. The great value of this potent vox populi, in this case vox det also, for fixing standard vocabulary and grammar cannot be over- estimated. It needs no mysterious telepathy to account for such phenomena. ‘They are grounded in the law of association, in the clan organism, and, since biological endurance is a fixed quantity, they ripen together. INDUSTRIES. The common and widespread interests in the activities of life, called industries, give rise to much simultaneity and identity of mental operations. Children go to school in common, the labor- ing class move to their employments as one, In the country they have a fashion of cutting a mark in the south kitchen window to note the noon hour, All housewives watch the 19044.) MASON—RIPENING OF THOUGHTS IN COMMON. 151 shadow of the window frame as it falls there and blow the dinner horn. You can imagine a wave of this joyful sound sweeping across the continent every day from ocean to ocean, and its pre- cisely similar effects on the spirits and bodies of millions on the farms, constituting an aggregate appetite. In precisely the same way the social and political life is agitated, and yet men are amazed to find themselves warming up on the same topics. In Washington City there are fifty thousand employees. They go to their work at a certain signal. Just at standard noon the whistles blow and they simultaneously and without consultation drop their work. There is a story going around of an old cabinet- maker in one of the Departments; who was so punctual in this regard that once when he was driving a nail and the whistle com- menced to blow, he left his hatchet up in the air, like Mohammed’s coffin, and went to his lunch. It is often said that women are governed by instinct, but men by reason. ‘The former share more thoughts in common, they are more conservative, even in savagery. So the actions performed over and over pass into semi-automatism, and without notice the thoughts associated with them arise together in many minds. Even the thoughts go in sets and cliques, and one will awaken the rest by association. Now and then in the industrial world, through the pleadings of environment, the inspiration of genius, the intense rivalries of trade, new tools, devices, processes and products, and new harness for the forces of nature are devised. The purely original in these are the exceptions, not the rule of action; and, besides, there is more survival than new creation in any one of them, as the suits for interference in the Patent Office will demonstrate. FINE ART. The esthetic faculty affords, with its schools and even national styles, most wonderful examples of the force of emotions felt in common. Canons of criticising the methods of appealing to the senses may be defined as expressions of the thoughts which artists of a certain epoch or school have come to hold in association. The same faculty becomes mixed with social life and gives rise to fash- ions and styles. Hence they say you might as well be out of the world as out of fashion. It will be asked whether this community among the agents and 152 MASON—RIPENING OF THOUGHTS IN COMMON. [April 9, agencies of enjoyment accounts for otherwise inexplicable concur- rences in art expression. The foundation of art, as of all other human actions, is laid in nature. That artists without consulting one another should copy this or that feature of the world around them is not surprising. But art is limited in execution. Tennyson’s prayer, “I would that my tongue could utter The thoughts that arise in me,” has been breathed by every artist that ever lived. Fatigued with failure he falls back on his fellow-workers, on the habits of the guild, on conventionalism, which is art-methods in common. It is wonderful how far and wide, and how long these survive. When a student of form in design, familiar with scrolls and frets in . Grecian art, discovers the same forms wrought out on Pima Indian basketry and lacework, he lifts his hands with surprise. The eth- nologist knows that the Indian woman has not necessarily held converse with the countrymen of Phidias. He realizes that the Pima woman is in the preparatory school, of which the Greek artists were full graduates. Once upon a time Grecian women wove into perishable basketry (zdvacrpa) forms that have never died and which their descendants fixed imperishably in marble. Besides the throng of specially endowed creators of art forms céoperating to their origin and perpetuation, there is a united, I almost said organized, admiration-in-common by the enjoyers or consumers of art products. Their habits of judgment, or canons, are intensified and fixed by custom. SociaL LIFE. The phrase ‘‘ social life’’ is here used in its most comprehensive sense, taking in the sources of all artificial activities performed by persons working unitedly. Two men managing a canoe down a rapid are intensely social; any rupture in the common thought would be fatal. Social organizations furnish the occasion for growth in what is here under consideration. They are like propa- gating gardens, farms, or stock ranges, where plants and animals are raised in vastly greater numbers than nature unaided would produce. 2 It would do no violence to partnerships, corporations, trades unions, and guilds in the industrial world ; to secret societies, clubs, 1904.) |MASON—RIPENING OF THOUGHTS IN COMMON. 158 and associations for cultivating the true, the beautiful, and the good, in the moral and intellectual world ; to the family, the clan, the tribe, the state, the nation, in the regulative world, with parlia- ments, courts, administrations, armies and navies, to characterize them as institutions for creating and preserving mental activities in common—popular legislatures that never adjourn. They afford also fields for their operation. When coincidences occur under their sway, the causes lie in the very nature of society from the beginning. It is an error to think that social structures and their demands become simplified as one descends from civilization, through barbar- ism, to savagery. ‘The abundant studies of Major Powell and his colleagues among the tribes of America, and of Morgan among savages in general, teach the contrary. Assuming that social structures and functions among these tribes are in the main types found in all primitive societies in the past, it is not difficult to understand how at the very outset the first society developed a vast number of thoughts in common that have persisted in all ages and areas. ‘To these must be added similar processes originating in races and smaller groups. Recall how immensely stronger are the character and marks of race than of individuals, how the latter vary in color, stature, via- bility, number and sex of children, and so on, But the race stature and number of births in males and females, as well as other characteristics of the species, endure. «So careful of the type she seems, So careless of the single life.” So long has nature moved by measured steps that there have come about not only cosmic thoughts in common, animal communis sensus, anthropic or human rhythm of mental action, but also rac- ial idiosyncrasies, national impulses, civic likenesses, industrial coincidences and inventions, and family likenesses. It is common to hear such expressions as ‘‘ the times are ripe,’’ ‘‘the hour has struck,’’ for this or that scheme, meaning that thoughts, like heads of wheat in summer time, have simultaneously ripened for the har- vest. It accounts for revivals, the singing of masses, the frenzy of crowds, and all such phenomena. The man of old who was wise enough to foresee these maturing of thoughts in common was recognized as a seer. If forceful enough he became a prophet, a leader, a reformer, a culture-hero. 154 MASON—RIPENING OF THOUGHTS IN COMMON. [April 9, LEARNING AND JLORE. Lore is the learning of the folk, the philosophy of savages, the survival of old beliefs and customs into enlightenment—old thoughts in common gone to seed. In form, it is the traditions, songs, proverbial philosophies, ceremonies, and real knowledge of peoples. The lore-thoughts of a people are the most deep-rooted and persis- tent, because indigenous to their minds. It is said that at the battle of Sebastopol the critical charge was incited by the play- ing of the Marseillaise, which the old soldiers heard for the first time in years. Anyone who has tried to oppose an absurd popu- lar belief, such as that in the hoop-snake, the retiring of the ground-hog at Candlemas, the marvelous doings of the earwig, and a thousand more, will appreciate what is here insisted on, namely, that the holding of a thought in common intensifies its activity, as in a battery of infinite number of cells. On the intellectual side, lore has become learning and ‘science is slowly permeating the communal mind and becoming the institu- tional mind. The personal equation of conservatism still acts as a balance-wheel there, as those who worked for uniform time and better nomenclature, and are now laboring for a uniform alphabet and a standard numeral system, will testify. The sciences began in the individual observation and were perfected one by one in the institutional mind. Since anthro- pology is a composite science, using and depending on all others, it will be the last to rise to the dignity of a perfected science. The same is true ofall its component sciences. In the museum one sees the botanist returning with his plants from the field. He has been collecting, he is a collector, these are his collections: his work is in the collective stage. Next, on long tables, he lays them in heaps, according to certain classific concepts in his mind, he is classifying, he is a systematist : this is his classification. Finally he comes to conclusions, will tell you beforehand what to look for in this or that class. He predicts, he is a philosophic botanist: his work is in the predictive stage. But this has been going on for centuries, with fresh returns to the fields, again and again with brighter eyes and larger experience. At last the organized mind takes up the task, so that the work of each must pass the scrutiny of all. 1904.] MASON—RIPENING OF THOUGHTS IN COMMON, 155 a RELIGION. So far as it enters the scientific arena, religion has to do with a spirit world and its influence on the world of sense. What is thought in common about that world, its physiography and its inhabitants, especially their activities among men and things, goes by the name of creed; what is done in common by men in the organization of society and in conduct responsive to creed is cu/t or worship. The most overpowering thoughts in common have belonged to the realm of religion. Things change and thoughts with them, not rapidly but surely. The unseen is not known to change, is believed not to change. The words of Paul, ‘‘ For the things which are seen are temporal ; but the things which are not seen are eternal,’’ embody a thought common to all races and ages, and have held all humanity in lines of conduct more firmly than the teachings of experience. To sum up, similar words and actions arise among men, spon- taneously and incessantly, not so much by reason of similar envi- ronments and provocations on the spur of the moment as from a psychological cause, the possession of thoughts in common that have come down through the ages and gathered velocity and impetus as they rolled. If subtle, telepathic influences exist in spiritual connections, they grow out of common thinking, they are the effect, not the cause, of striking coincidences. To the educator, the reformer, and the legislator, no less than to the investigator, a constant realization of this fact is necessary to success. To those who listened to this paper, necessarily brief and general, multitudes of instances will arise where strange coincidences in conduct have expressed themselves in every line of activity. If they were not too busily engaged with the affairs of life they would have noticed many more; because with a normally constituted mind and in a completely organized society they are the rule and not the exception. 156 OSBORN—THE EVOLUTION OF THE HORSE. [April7, RECENT ADVANCES IN OUR KNOWLEDGE OF THE EVOLUTION OF THE HORSE. BY HENRY F. OSBORN. (Read April 7, 1904.) The American Museum explorations for the development of the horse practically began in tgor with the first expedition to the Rocky Mountain region in that year, conducted by Dr. J. L. Wortman. By continued exploration and the acquisition of the Cope Collection of fossil vertebrates remains of a large number of fossil horses were secured. In 1go1, however, explorations were organized with the particular purpose of securing materials for the further study of the evolution of the horse with the fund donated by the late William C. Whitney. Mr. J. W. Gidley, a graduate of Princeton University, was placed in charge of expeditions sent into Texas, Colorado, South Dakota and Nebraska. The remains of 146 horses were secured, making a total of remains representing this animal in the Museum of upwards of 770. In the year 1990 the chief discovery was a herd of six Pleistocene horses belonging to the new species Lguus scott’, giving us for the first time a complete knowledge of the osteology of the American Pleistocene horse—a large-headed, short-limbed animal, propor- tioned somewhat like the zebra. In 1901, the first year of the Whitney expeditions, Hypohif~pus was discovered in the Upper Miocene, a genus named by Joseph Leidy but hitherto little under- stood ; this animal, although contemporaneous with several highly specialized types of horses, was found to represent a forest-living type, with short crowned teeth and persistent lateral toes. In 1902 the remarkable discovery was made of a new genus and species of horse, Meohipparion whitneyi, in the Upper Miocene of western Nebraska. This animal, in contrast with the foregoing, was extremely light limbed, proportioned rather like the deer, with diminutive lateral toes, long crowned teeth, and represented a highly specialized, cursorial type, remotely related to the A/ippar- jon of Europe. Our explorations therefore have demonstrated the existence of two and probably three collateral lines of horses contemporaneous with the Protohippus line, which apparently led into the true horse, The early conclusions of Joseph Leidy, based on far less 1904, PHILLIPS—RADIUM IN AN AMERICAN ORE. 157 perfect material, are thus confirmed in the most gratifying manner. Many problems yet remain to be solved, however, especially the osteology of the line leading directly into the modern horse. Explorations will therefore be continued, especially the search for the skeleton of Protohippus, with a view to ascertaining whether this is or is not one of the direct ancestors of Zguus caballus. American Museum of Natural History, New York, April 7, 1904. RADIUM IN AN AMERICAN ORE. BY ALEXANDER H. PHILLIPS. (Read April 8, 1904.) The work which. I have accomplished in the separation of radium, or more exactly the concentration of radium in barium salts, has been carried on entirely with the mineral carnotite. Carnotite is comparatively a new mineral, having been described by Friedel and Cumenge in July, 1899, and for this reason it is not found in most books on mineralogy, and is therefore but little known to the general prospector. It was first discovered in the western part of Colorado, and occurs in Montrose, San Miguel and Mesa counties of that State and the adjacent counties of Utah. The theoretical percentage composition as given by Friedel and Cumenge is: eR Ls ine niltnin'e plata ance p ovacuia kx eihiace td GoMR aS aR a hoa 63.54% RRR o's ai5:4:sip ig. av Hay Cake < ta 9.a ooh WAS RRR AT 20,12 * OTS 6 6 05 os pO bass Cys nny 0 4s ee ee ee eee 10.37 “ PUM Ss Sanne Gee esas ea yet ds aan mesa Tey OeeaeD 5-95 * Results very close to these were obtained in the actual analyses. The mineral formula is given as 2UQ,, V,O,;, K,0O, 3H,O, or a uranyl potassium vanadate with three molecules of water of crystal- lization. Hillebrand, after a series of analyses, disputes this com- position, and holds that the mineral is probably a mixture to which the above simple formula is not applicable. Carnotite is a light canary-colored powder disseminated through a fine-grain sandstone. It is easily soluble in acids, and is treated in this way for the commercial production of uranium salts. 158 PHILLIPS—-RADIUM IN AN AMERICAN ORE. [April 8, In October, 1902, I received twenty-five pounds of this ore from Richardson, Utah. Carnotite occurs here in the usual way; the ore being rather a lean one, no specimens of which carries more than 10% of the mineral carnotite, while the average is greatly below this amount. The percentage content of uranium and vanadium in this ore was not determined, which is to be regretted. The radio-activity compared to uranium, as determined by G. B. Pegram, of Columbia University, was .40. After a series of experiments upon a one-pound sample, the remainder of the twenty-five pounds was treated as follows: It was first leached with hydrochloric acid to remove most of the soluble salts ; as radium salts are isomorphous with barium salts, and agree very closely in their chemical properties and solubilities with the similar barium salts, it was thought that strong nitric acid would dissolve the small amount of barium the ore contained and also the radium with it. After the hydrochloric acid treatment the insoluble residue was treated with concentrated nitric; these two acid solutions were concentrated to small volume. Upon testing them for barium it was found that the precipitate would be small, so it was thought advisable to add a small amount of barium chloride, which would act as a carrier and help in the separation of the small amount of radium present; sulphuric acid was then added; the solutions diluted and allowed to stand several days to settle; the clear solu- tions siphoned off. The resulting sulphates after washing were fused with alkali carbonates. The melt dissolved in water, the insoluble carbonates after washing free of sulphuric acid were dis- solved in hydrochloric acid. Hydrogen sulphide was then passed through the solution to free it of lead. From this solution free of lead the barium was precipitated as a carbonate, and dis- solved in the least possible quantity of hydrochloric acid, when the barium and radium will be in the form of chlorides and are in a condition to concentrate the radium by fractional crystallization. This solution of chlorides was allowed to evaporate slowly until about one-half of the contained salts had separated as crystals, when the crystals were removed, redissolved, the solution again allowed to evaporate slowly until one-half had separated, and so on for a third time, when there was obtained as a final product a little less than one-half gramme of chlorides, the activity of which com. pared to uranyl nitrate was about 1500. 1904.] PHILLIPS—-RADIUM IN AN AMERICAN ORE. 159 The residual chlorides amounting to two grammes were recovered, and proved to be quite active also; their activity as compared to uranium was measured by Pegram as 365. The radio-activity of the final product could be increased by several fractional crystallizations, when a specimen much less in weight, but more active, would be obtained. The radio-activity of the specimen as obtained was deemed sufficiently high to indicate that radium could be produced in quantity from carnotite, at least from this locality, as twenty-five pounds of rather a lean ore had been used. Had aton been worked over in the same way it would yield a gramme of chlorides of 60,000 radio-activity as compared to uranium. This specimen was sepa- rated in November, 1902, and is as active now as then. This establishes without doubt the fact that radium salts are’ dissolved in the acid with which the uranium minerals are treated in the commercial preparation of uranfum salts. In the crystal- lization of these salts the radium would be carried in connection with the uranium, as it is in the natural formation of the uranium minerals. This would explain to a certain extent the variable activity of uranium salts, as their activity is not proportional generally to the uranium which they contain. Since the separation of this first sample of radium from carnotite, I have received specimens of the mineral from other localities, all of which are active, their activity depending upon the amount of carnotite in the ore. One specimen of quite pure mineral gave an activity of 4, the highest observed in the crude ore. A short time ago 3.5 kilos (about eight pounds) were obtained from Montrose County, Colo., selected specimens of which were exposed, in the ordinary way, in making X-ray negatives, with very satisfactory results. These photographs show very clearly the bands of active carno- tite separated by the inactive matrix. The plates used were Carbutt’s B, and exposed to the action of the mineral for sixty hours. With more sensitive plates the same effect could be obtained in much shorter time. I am at present at work upon the separation of the uranium, vanadium, and radium salts from these eight pounds of ore as an exhibit at the St. Louis Fair. This work is as yet not completed. After pulverizing and thoroughly mixing, its radio-activity was measured as 1.71, compared to uranyl nitrate. It was then treated 160 PHILLIPS—RADIUM IN AN AMERICAN ORE. [April 8, with dilute acids, as it was intended in this case to boil the insolu- ble residue in a concentrated solution of sodium carbonate, to extract the last of the radium. After the acid treatment the residue from the 3500 gms. originally taken weighed 2200 gms. and gave ‘an activity of 1.40. The solution contained 1300 gms. of the amount taken. If the activity is calculated to gramme units compared to uranyl nitrate the 3500 gms., activity 1.71 = 5985 units. Residue insoluble in dilute acids, 2200 gms., activity 1.40 = 3080 units. The solution contained, therefore, 2905 units, considerable of which would be due to the uranium dissolved and emanation from the radium. The dilute acid extracted nearly all of the uranium, but there was still some found in the residue. This ore contained con- siderable barium and a large amount of calcium. No barium was therefore added to the solutions, as before. After precipitation of the sulphates, and the separation of barium from other bases, 3.8 gms. of barium carbonate were obtained, which gave an activity of 35-8, or 135 units compared to uranium nitrate, ‘This activity was measured upon the same day (#.e., yesterday) that the barium salt was separated from the cther bases; it may be expected that its activity will increase, from the accumulation of the emanation ; this increase may in some cases amount to several times the original activity of the compound when first prepared. This demonstrates that dilute, acids, while they dissolve consider- able of the barium salts present, the greater proportion of the radium is still left in the residue ; but even this amount, which is small com- pared to that left in the residue, as indicated by the radio-activity of the residue, if calculated upon the basis of a ton, would yield a gramme of chlorides of 11,300 activity as compared to uranium. These facts prove beyond question that carnotite will become a commercial source of radium. Princeton University, March 7, 1904. 1904, , ABBOTT—ARTIFACTS BENEATH DEPOSIT OF CLAY. 161 ON THE OCCURRENCE OF ARTIFACTS BENEATH A DEPOSIT OF CLAY. BY DR. CHARLES CONRAD ABBOTT. (Read April 8, 1904.) A recent examination of the surface soils of a shallow valley-like depression in upland fields, elevated 50-70 feet above the Delaware River and its flood-plain, made evident that the present little brook was not the original and only watershed of this tract of land, but the remnant of a stream of greater volume which had at one time practically filled the valley. ‘To reach the flood-plain of the present river, the brook of to-day passes through a deep valley that has been worn into the face of the bluff that extends for along distance parallel to the river’s course. A bird’s-eye view of the region shows at a glance that when the present. flood-plain was permanently under water, the gully did not exist in its present width and depth and the greater volume of the present brook emptied directly into the river. As the river’s volume decreased and the stream con- fined itself to the channel now existing, the brook wore away the face of the bluff until it reached the abandoned river bed or what is now a wide meadow, ordinarily dry and cultivable, but occasion- ally overflowed to a considerable depth. In a cross-section of the upland valley, extending over two hundred feet in width, it was found that immediately below the present soil and deeper sand as yet unaffected by decomposition of vegetable matter, there was a well-defined deposit of clean, sharp river sand, a few pebbles and a trace of clay that resulted in a slight cementation of the mass. Besides this condition, there was at one part of the section, some forty feet in extent, a deposit of clay, comparatively free from grit and so compact that no object could have intruded. It was nine inches in thickness and twenty-four inches from the surface of the ground to its base, which was com- pact coarse sand, pebbles and a little clay. Resting on this base, an unquestionable bed of a water-course, were artifacts, consisting of flakes of argillite, artificially produced and the hammer-stones, or oval pebbles, with the ends battered by continued violent contact with other minerals. A closer examination of the spot indicated clearly that the clay PROC. AMER. PHILOS. SOC. XLII. 176. K. PRINTED JULY 13, 1904, 162 ABBOTT—ARTIFACTS BENEATH DEPOSIT OF CLAY. [April 8, was derived from beds of Raritan clay, near the head of the valley. For some reason, not now definable, this clay, taken up from the bed of original deposition, was redeposited in a circumscribed area and, as it proved, at a point where previously traces of man had been lost in the waters of a prehistoric stream. Later, this upland stream had decreased in volume, shrinking to the present trifling brook. The bed of the greater stream had become choked with vegetation and eolian sands had drifted in until the valley was no longer well defined ;. its one-time features finally disappearing when forest trees for centuries thickly dotted the ground. For how long this condition continued it is impossible to determine. When the valley was a forested tract the Indian was in possession. For somewhat more than two centuries the forest has been gone and the ground under more or less constant cultivation, but the valley is still to be traced. In the mid-autumn of 1903 there was a phenomenal flood in the Delaware River. The water rose to a height unrecorded by man, and the river reasserted its right to the flood-plain and beyond, for its waters flowed up the ravine and the hillside brook became for the time a navigable stream for a considerable distance. It was clearly a return for a time to those ancient days when the entire condition of the country was essentially different from what now obtains—when little brooks were considerable creeks, when creeks were rivers, and the river itself a stream that approached the present Mississippi in magnitude; and when this was true of the tamer conditions of to-day, not only man but an arctic fauna lived here. We can only refer such conditions to the closing days of the Glacial Epoch. The question of Glacial Man in North America, so long a vexing problem, has been found easier of solution than was to be hoped for. The evidence above given is nota single instance of deeply inhumed artifacts in undisturbed stratified deposits; but is submitted as a typical example of many such that have been brought to light by the author and other workers in the same field, Trenton, N. J., April 7, 1904. 1904. | ABBOTT—REPORTED SHOWERS OF TOADS. 163 ONE EXPLANATION OF REPORTED SHOWERS OF TOADS. BY DR. CHARLES CONRAD ABBOTT. (Read April 8, 1904.) The frequent references in newspapers to occurrences of ‘‘ showers of toads ’’ have suggested to the author that a condition in the life- history of the spade-foot toad, a little-known and strictly nocturnal species, living in the ground, might explain them more rationally than that the little batrachians are picked up by the wind in one place and dropped in another, perhaps miles away, or that other still more strange view quite common among the ignorant that toad-spawn is sucked up by the sun and hatched in clouds, where the tadpoles remain until they have advanced to the dignity of hoppers, when they fall to the earth. Unlike the common toad and the frogs, the spade-foot toad (Scaphiopus solitarius) does not have a regular season for deposition of ova, but the eggs may be laid at any time from April 1 to August 31. Furthermore, this batrachian does not resort to permanent watercourses or ponds on such errand, but takes advantage of temporary pools formed by showers of longer duration than is usual. It is remarkable how admirably this strange irregularity of an important event should be adapted to transitory conditions. Pools of rainwater seldom remain long on the ground’s surface. Soakage and evaporation soon obliterate them; but that this may not prove a fatal objection, the eggs of the spade-foot toad hatch in about ninety-six hours, and in less than two weeks, or fourteen days at most, the tadpole has become a terrestrial animal or a ‘‘hopper’’ and leaves its nursery. The development is even more rapid occasionally, I am led to believe, being accelerated by excessive warmth or retarded if the days are cool and cloudy. It will be readily seen that young spade-foot toads, congregated in or immediately about a temporary pool, will not wander far from it when their subterranean life begins, but will bury themselves in the comparatively moist ground where they happen to be. Should, at this time of their limited wandering, there occur one or more violent showers, the ground being wetted and little pools formed, the young spade-foot toads would necessarily, we might say, wander over a much wider extent of territory, and, escaping notice when 164 LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. [April 7, confined to one fast disappearing pool, would be observed when dotting the ground over an extent perhaps of an acre or more. Seen thus, immediately after rain, and not previously noticed, the inference is not so strange that they came to the earth with the rain, or that there had been a shower of toads as well as of water. Trenton, N. J., April 7, 1904. EXPANSIONS OF ALGEBRAIC FUNCTIONS AT SINGULAR POINTS. BY PRESTON A. LAMBERT. (Read April 7, 1904.) I. INTRODUCTION. An algebraic equation F(x, y) =o of degree in y defines y as an m-valued algebraic function of x. When these x values of y are all distinct for a given value of x, that value of x is called a regular point of the algebraic function, and the # branches of the function are extended by applying the law of the continuity of each branch. In curve tracing x and y are real variables and only the real branches of the function are used. Real values of x and y which satisfy the equations /(x, y) = 0, oa =O, > == o determine mul- tiple points of the curve which represents the equation F(x, y) =o. If «=a, y=4 isa multiple point of this curve, the behavior of the curve at the multiple point is determined from the expansions of y— in terms of «—a. Inasmuch as the transformations =H +a,y=y,+6 transfer the origin to the multiple point, the multiple point will always be taken at the origin. An algebraic equation between complex variables F(w, 2) =o of degree ” in w defines w as an n-valued algebraic function of z. Values of w and « which satisfy the equations A(w, z) =o and — F(w,%) ==0, determine branch points of the algebraic func- tion, that is points where several branches of the function meet, The behavior of the function at a branch point is determined from the expansions of the function at the branch point. 1904] LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. 165 The multiple points in curve tracing and the branch points in algebraic equations between complex variables are grouped as the singular points of algebraic functions. II. HisTorIcau. , The problem of the expansion of algebraic functions at singular points dates back to Newton. Newton’s ‘‘ parallelogram method ”’ determines the first term of the expansions as follows. The equa- tion transformed to the singular point as origin becomes 2 Gan?" Y" = 0. Locate on squared paper to rectangular axes the points (m, n) whose codrdinates are the exponents of x and yin the various terms of the transformed equation. Connect by successive straight lines, forming a broken line convex toward the origin, the points nearest the origin. The sums of the terms of 2a,,,x"y" = 0, for which the points (m, #) are located on the same straight line, when equated to zero form equations which determine the first terms of the expansions at the singular point. Puiseux in his classical ‘‘ Memoir!on Algebraic Functions,” Liouville’s /Journa/, t. XV, 1850, used Newton’s parallelogram method and studied in detail the nature of the expansions of algebraic functions. Puiseux’s Memoir is made the basis of Briot and Bouquet’s ‘‘ Elliptic Functions,’’ and indeed is almost univer- sally used in the study of algebraic functions. N6ther’s method in Annalen, 1X, 1876, is representative of the more recent methods of expansion of algebraic functions. By suc- cessive quadratic transformations the singular point becomes a regular point, and from the expansions at this regular point the expansions at the original singular point are obtained by reversing the quadratic transformations and the reversion of series. In the present paper an analytic method is presented which determines not only the first terms of the expansions but also the successive approximations of the several expansions. The method of expansion used is that application of Maclaurin’s series which the author employed to compute all the roots of numerical equa- tions and which is published in Vol. XLII of the Proceedings of the American Philosophical Society. Ill. A New Metruop or EXPANSION. For convenience of description the exponents in the equation to which the method is applied are assumed numerical. 166 LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. [April 7, Suppose the algebraic equation when the singular point is taken as‘origin to have the form (1) Gaty + Faty® + Hay? + Jaty" 4 Koby” + Ey + Lay? + Daty’ + Lay! + Cxly + Bay + Ax —o, . where the terms are arranged according to the descending powers of y. In this equation y has fourteen branches, which are to be separated at the singular point by expanding y as a function of x. The terms of equation (1) to be underscored are determined by the method used for this purpose in the paper on the ‘Solution of Equations,’’ and which is adapted to the present case as follows. Tf Lamy, Mx™ ym, Mx™sy™s are any three terms of equate (1), the value of (2) limit Ma,—o5 x™2(21—23) eo Less AAs | eB a— 8) 483815) is zero, finite, or infinite. It is at once seen that this limit is zero, finite, or infinite, according as m,(”,— m,) is greater than, equal to, or less than ,(7, — ”,) -+ m,(”,— m). The underscored terms of equation (1) are all the terms which satisfy the following condition. The limit (2) for any three consecutive underscored single terms is infinite. If a group of terms is underscored as a single term, the limit (2) is finite for all the terms of this group, and the limit is infinite for the first term of the group and the next preceding underscored term, the limit is also infinite for the last term of the group and the next succeeding underscored term. We now proceed to underscore the terms of equation (1) to satisfy this condition. Underscore the first term of (1), and determine the limit (2) for the first three terms of (1). Since the limit is infinite, underscore the second term of (1) temporarily. Determine the limit (2) for the terms 2, 3, 4. The limit is zero and term 3 is not underscored. Next determine the limit (2) for the terms 2, 4, 5. The limit is infinite and term 4 is tempor- arily, term 2 permanently underscored. The limit for terms. 4, 5, 6 is zero, the limit for terms 1, 2, 6 is infinite. Hence term!4 does not remain underscored, and term 6 is temporarily underscored. 1904.) LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. 167 The limit for the terms 2, 6, 7 is infinite, the limit for terms 6, 7, 8 is zero, the limit for the terms 2, 7, 8 is infinite. Hence term 6 permanently underscored, term 7 is not underscored, and term 8 is temporarily underscored. ‘ The limit for terms 6, 8, 9 is infinite, for terms 8, 9, 10 zero, for terms 6, 8, 10 infinite. Hence term 8 is permanently under- scored, term 9g is not underscored, and term 10 is temporarily underscored. The limit for terms 8, 10, 11 is infinite, hence term to is permanently underscored. The limit for terms 10, 11, 12 is finite, and these three terms are underscored as one term. The several equations formed by retaining in equation (1) in succession only consecutive underscored terms, if these terms are single, and if a group of terms is underscored by retaining only the group of terms, the first term of the group and the next preceding underscored term, and the last term of the group and the next succeeding underscored term, will determine the first approxima- tions of the fourteen branches of the function. These equations are a) Gx*ty+ F=0, 6) Fe*y’+EF=0, c) £¥+ Dx*=0, ad) Dy’ + Cxt=0, ¢) GO + Bx*y+ Ax‘ =0, and the fourteen first approximations are 4 a) 9=—24, dy=(-4) 5 O9=(-$) a a y=(—£) A, 97 =~ ABV Of these fourteen branches the separate branch @) and the three separate branches 4) go through infinity when x0. The cycle of five branches ¢), the cycle of three branches @), and the two separate branches ¢) constitute the ten branches which meet at the singular point. If a factor ¢ is introduced in succession into all the terms of equation (1) except the terms used to determine the first approxi- mation of a branch of the function, the successive approximations of this branch are determined by developing y in ascending powers of t, x considered constant, by Maclaurin’s Series and making ¢ unity in the result. 168 LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. [April 7, IV. BEHAVIOR OF BRANCHES OF ALGEBRAIC FUNCTIONS. If the algebraic equation takes the form 2a,,x*"y° =o when a regular point is taken as origin the method of expansion deter- mines # separate branches of the function. If the origin is a singular point of Sa,,,x*™y" —o the behavior of the several branches is determined as follows. a) If the first approximation is independent of x or contains a negative power of x, the corresponding branches are either finite or infinite and consequently these branches do not go through the singular point. 4) To each pair of consecutive underscored terms in which the exponents of y differ by unity there corresponds a separate branch of the function through the singular point. ¢) To each pair of consecutive underscored terms in which the exponents of y differ by more than unity there corresponds a sepa- rate cycle of branches hanging together at the singular point, pro- vided the exponents of x and y in the equation determining the first approximation are prime to each other, and the number of branches in the cycle equals the exponent of y. If, however, the exponents of x and y in this equation have a common divisor greater than unity, the corresponding branches break up into cycles equal in number to the common divisor and the number of branches in each cycle is the exponent of y divided by the common divisor. d) If a group of terms is underscored and the equation formed by equating this group to zero has equal roots, these equal roots must be removed before the branches corresponding to the group can be separated, If this equation is now solved the branches will be separated into single branches and cycles of branches, provided the exponents of y in this equation have no common divisor greater than unity. If, however, there is a common divisor greater than unity the branches corresponding to this group break up into sub-cycles. V. APPLICATIONS IN CURVE TRACING. Example J.—Let it be required to trace the curve represented by the equation (1) yy —gaty — xty + gx'y + 2x° —2x" = 0 in the neighborhood of the singular point (0, 0). 1904.] LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. 169 Collecting terms in like powers of y (2) ¥° + (— 3x*— 2° + 92x") y + (2x° — 2x") = 0. Since in a’ first approximation the lowest powers of x in the several coefficients alone count, this equation may be written (3) 9° — 3x4y + 2x° = 0, The application of the method of underscored terms shows that in equation (3) the three terms must be underscored as one term, hence (4) 9° — 3x'y + 22* <0. The equation y*— 3x‘y + 2x*—=o has two roots each equal to x? Diminishing each root of equation (1) by x’, if we write y= Ji + x’, we obtain the equation, (5) Ja? + 3491" + (— 2° + 92") ny + (— 32° + 9x") = 0. Retaining for a first approximation only the lowest powers of x (6) n° + 3xiyy? — a'y, — 32° = 0. In equation (6) the terms 1, 2, 4 must be underscored, that is (7) ri + 3x%y.? — x*y, — 32° = 0. From equation (7) the first approximations of y, are (8) A= — 3%, =, y= — *. Consequently the first approximations of y are (9) y= ae, y= att ot, yaa — a, The three branches of y are separated by these approximations and the behavior of the curve at the multiple point is found by tracing the three equations (9) in the neighborhood of the origin. Example JI.—Let it be required to trace the curve represented by the equation (1) ° —3xy + 9x'y + 2x° = 0 in the neighborhood of the singular point (0, 0). 170 LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. [April 7, To obtain a first approximation this equation may be written (2) »— 3a"y + 2x 0. The three terms of this equation must be underscored as a single term, when it is found that the equation from which the first approximations are to be found has two roots each equal to x’. Transforming equation (2) by writing y = y, + x’, there results (3) Jv? + 3%’ + 9% + 9%? = 0. In equation (3) terms 1, 2, 4 must be underscored, which gives (4) wet 3x7y,.? + oxy, a gx? =0. The first approximations of y, are (5) zA=— 3%", ¥, = 3h", J, = — Bis", and consequently the first approximations of » (6) y= — 20%, y = + ict, y = 2? — gin! The approximations (6) separate the three branches of the curve at the multiple point. VI. APPLICATIONS IN FUNCTIONS OF THE COMPLEX VARIABLE. Example .—Let it be required to determine the behavior of the five-valued algebraic function defined by the equation () #@—(1—#) wt — 4 (1x—2)'=0 at the branch-points of the function. The branch-points, the common solutions of (1) and the partial derivative of (1) with respect to w, (2) 5w'—4 (1 —2') w=0 are located at z==0, s== +1. At s==o the first approximations of w are determined by the equation 1904.) LAMBERT—EXPANSIONS OF ALGEBRAIC FUNCTIONS. 171 These first approximations are 3 2 2 z I, az’, — az”, aiz’, —aiz’, where a satisfies the equation 4 4* a> . 5° This shows that at the origin there is one separate branch, and two separate cycles of two branches each. To determine the behavior of the function at z= + 1, place z = #+1/in equation (1). There results 4* 5 (4) w — (F 2c — 2") wt — 8 (2' F 1)? (F 22 — 7 )*=0, which for a first approximation may be written 6 4 (5) w+ ad’ w* — = zs’ =0. The first approximations are Wf = = (4)* 2! from which it is seen that at the branch-points z = + 1 five branches of the function hang together in a cycle. To determine the behavior of the function at the point z= o, w == o, substitute in (1)z = =) w= =; whence (6) $2") w" 2" 2") w — =o. Equation (6) for a first approximation at («z= o0, 7 =o) may be written Equation (7) has two roots each equal to 22". Increasing each of the five roots of (6) by + rad * and retaining for a first approxi- mation only the lowest powers of 2’ in the several coefficients, 4 3 2 e (8) Safe wl 4 Saw 4 ea 2 af 2" =o. Equation (8) shows that at the point (w= 0, z = oo) the func- tion has five separate branches, that-is the point at infinity is not a branch-point of the algebraic function. 172 LAMBERT—EXPANSIONS OF ALGEBRAIC FUNC'LIONS. [April 7, Example II,—To illustrate the method of finding the successive approximations let it be required to determine to three terms the expansion of the branches-of the cycle corresponding to the under- scored terms of the equation (1) w — sta* —s'w—2" =0. Introducing a factor ¢ into the terms of (1) which are not under- scored, then differentiating twice with respect to ¢ COneehS 2 constant, . w — su? — swt — f= 0: a (3) 5a — area chante 3 ext'0, (4) (5* — 22* w— st) imptena!: a — 28 FZ =. Making ¢ =o in (2), (3), (4) $ ;dw 3 ys (Pw ** (w), = (sph =t8 + $3 (Ta h=—Fs . Substituting in Maclaurin’s series 2 w= (w+ (Gee t+ (Gale ts + - and making ¢= 1 in the result we find (5) w= st + 4patt geY which is correct to three terms. Equation (5) has the form of a power series in Pa beginning with the fourth power and represents a cycle of three branches of the algebraic function w hanging together at the singular point. Lehigh University, South Bethlehem, Pa., April 7, 1904. 1904.] JASTROW—THE HAMITES AND SEMITES. 1738 THE HAMITES AND SEMITES IN THE TENTH CHAP- TER OF GENESIS. BY MORRIS JASTROW, JR. (Read April 4, 19038.) & The roth chapter of Genesis is generally admitted to be one of the most remarkable but also one of the most puzzling documents of antiquity. Scholars have been engaged ever since the days of the Talmud’ and of Eusebius in attempts to identify the nations named in the chapter and in endeavors to determine the point of view from which the division of nations has been made and to ascertain the character of the underlying ethnological and ethno- graphical scheme, if there be one in the chapter. Modern research, aided to a certain extent by ancient tradition, has succeeded in identifying a large number of the nations enumerated,’ but the attempts to discover any system in the groupivg of the nations have failed chiefly because of the erroneous assumption that an ancient document could give evidence either of scientific accuracy or of ethnological finesse. An adequate conception of what really con- stituted a nation lay beyond the mental horizon of the ancient For a partial bibliography see Dillmann’s Gemess (Engl. transl. of sixth ed., Edinburgh, 1897), p. 325. For the Talmudical views and identifications see Neubauer, La Geographie du Talmud (Paris, 1868), pp. 421-429. 2See for recent expositions the commentarizs of Gunkel (1901), Holzinger (1899), Strack (1894), and Driver (1903) to the chapter in question; also Schrader, Cuneiform Inscriptions and the Old Testament (London, 1885), Vol. i, pp. 61-103; and Glaser, Skizse der Geographie und Geschichte Arabiens, ii (Berlin, 1890), chaps. 26 to 28. The chapter in Alfred Jeremias’ Das Alte Testament im Lichte des Alten Orients (Leipzig, 1904), pp. 145-170, is to be especially recommended as the latest summary of accepted identifications and because of its valuable supplementary statements, and suggestions toward the solution of the many problems in the 10th chapter of Genesis. A serious defect, however, of Jeremias’ treatment of the chapter is his failure to take sufficiently into account its composite character, consisting, as it does, of two distinct docu- ments together with many glosses and insertions, Thus, what he says about the supposed “ Arabian” origin of Nimrod (p. 158) falls to the ground if verses 8-12 are recognized as an addition that stands in no connection with verse 7; nor does Jeremias’ general view of the Vélkertafel as a unit (p. 145) commend itself in the light of the critical analysis of the chapter. 174 JASTROW—THE HAMITES AND SEMITES. [April 4, world, certainly of the ancient Orient. Apart from a certain instinct—to speak indefinitely—which correctly led a people to predicate its own closer or remoter relationship to others, reliance was placed on more or less uncertain traditions, and the value of such tradition was still further diminished by the subjective factors—a people’s likes and dislikes, its experiences and ambitions—that entered as elements into its formation; and when we pass beyond the immediate political environment of an ancient people, we must be prepared for a nebulousness of views that is almost inconceiv- able to a modern mind and for inconsistencies that are as bewil- dering as they are numerous. In view of this, it is evident that the critical analysis of the chapter to which modern scholarship has devoted itself with marked success is insufficient for a solution of the problems involved unless it also takes into account the uncriti- cal attitude of the ancient world toward ethnological and geo- graphical data. The critical analysis of the 1oth chapter of Genesis has reached a stage that may, with reasonable certainty, be regarded as definite and as having attained its utmost limits.‘ Of the two documents combined to form the present Vé/kertafe/—to use the convenient German term—the one that forms part of the Priestly Code, dis- tinguished by critics as P, forms the chief element, as is the case throughout the first eleven chapters of Genesis,’ while the other, designated as J, has only been drawn upon to the extent of furnish- ing supplementary data, though at times those supplementary data exceed in length the account in P, and, occasionally, J furnishes material, like the story of Cain and Abel, not found in P at all. In the case of the 1oth chapter, while J is actually longer than P, yet the latter document represents a far closer approach to a syste- matic arrangement, whereas J, marked by many glosses, is extracted in so arbitrary a fashion in order to supplement P that it is difficult to obtain an accurate view of the system followed by the “J” document in its original form, When the two documents are placed side by side, the differences between them will become clear. ? Wellhausen, Composition des Hexateuchs (34 ed., 1899, pp. 4-7): *Budde’s Urgeschivhte, pp. 499 sg. pp. 464-465, and also pp. 521-531, where the Jahwistic source in the first eleven chapters of Genesis is put together, 1903.] FP. 10, (1a) These are the generations of the sons of Noah; Shem, Ham and Japheth. (2) The sons of Japheth were Gomer, Gog,! Madai, Javan, Tu- bal, Meshech and Tiras. (3) The sons of Gomer were Ashkenaz, Rip. hath and Togarmah. (4) The sons of Javan were Elishah, Tarshish, Kittim and Rodanim,? (5) Of these the islands of the nations branched off. [These are the sons of Japheth] according t their lands, each according to his lan. guage, according to their clans among their nations, (6) The sons of Ham were Cush, Mizraim, Put and Canaan, (7) The sons of Cush were Seba, Havi- lah, Sabtah, Raamah and Sabtechah. The sons of Raamah were Sheba and Dedan. (20) These are the sons of Ham according to their clans, their language, according to their lands among their nations, (22) The sons of Shem were Elam, Asshur, Arpach- shad, Lud and Aram. (23) The sons of Aram were Uz, Hul, Gether and Mash, (31) These are the sons of Shem according to their clans, their language, according to their lands, ac- cording to their nations, (32) These are the clans of the sons of Noah ac- cording to their generations, among their nations and from them the nations we divided in the earth after the 1 Hebrew text has magog, which, however, appears to be an error for fog. *So read according to I Chr, 1, 6 instead of Dodanim. JASTROW—THE HAMITES AND SEMITES. 175 J. 9, (18) The sons of Noah that went forth of the ark were Shem, Ham and Japheth (gloss: ana Ham is the father of Canaan). (19) These three were the sons of Noah; and of them was the whole earth overspread... . . (10, ” to them sons were born after the . @) Cush begat Nimrod. [He was the rst mighty one in the earth, $9) He was a mighty hunter before Yah- weh: wherefore the saying: A mighty hunter like Nimrod before Jahweh. (10) The beginning of his kingdom was Babel, Erech, Accad and Calneh in the land of Shinar. (11) Out of that land he went forth to Assyria and founded Nineveh, Rehoboth-Ir, Calah and (12) Resen (between Nineveh and Calah) (gloss : that is the great city)). . «++ (13) Mizraim begat Ludim, Anamim, Lehabim, Naphtuhim, (14) Pathrusim, Casluhim and Capthorim (gloss: whence went forth the Philis- tines'). [( 15) Canaan begat Sidon his first born and Heth (16) (gloss: the Febusite, Amorite, Girgasite, (17) Hivite, Arkite, Sinite, (18) Ar- vadite, Zemarite and Hamathite))}, Afterwards the clans of the Canaanite spread, (19) so that the boundary of the Canaanite extended from Sidon to Gerar [gloss: to Gaza] to Sodom and Gomorrah [ gloss: to Admah and Ze- boim] to Lasha..... (21) And to Shem also (sons) were born, the father of all the sons of Eber, the elder brother of Japheth..... (24) Arpachshad begat Shelah and Shelah begat Eber- (25) To Eber two sons were born, the name of the one was Peleg (gloss: Sor in his days the earth was diviaed*) and the name of his brother was Joktan. (26) Joktan begat Almodad, Sheleph, Hazarmaveth, Jerah, (27 ) Hadoram, Uzal, Diklah, (28) Obal, Abimael, Sheba, (29) Ophir, Havilah and Jobab. All these were the sons of Joktan. (30) Their settlement was from Mesha to Sephar, the mountain of the east, 1Cf, Amos 9, 7; Jer. 47, 4; Deut. a, 2% 2 niphilega. 176 JASTROW—THE HAMITES AND SEMITES. [April 4, II. Beginning with ro, 1°, as a heading, ‘« These are the generations of the sons of Noah; Shem, Ham and Japheth,”’ P furnishes (verses 2-5) the list of nations sprung from Japheth and then takes up (vv. 6-7) the second son Ham, the close of which enumeration is to be sought in v. 20. Thirdly, Shem, the oldest son, is taken up (vv. 22-23), the continuation appearing in v. 31, while v. 32 represents the conclusion of the version as follows : ** These are the clans of the sons of Noah, according to their generations, according to their tribes and from them the nations were divided in the earth after the flood.”’ . It will be observed that in this compact survey, resting on the theory of the descent of all the nations of the earth from a single ancestor, Noah, through three groups represented by Noah’s three sons, there are decided inequalities in treatment. Of the sons of Japheth, only two, Gomer and Javan, are carried down into further subdivisions. In the genealogy of Ham, only one, Cush, is singled out for further subdivision, but this one is carried down through its branch, Raamah, into a further subdivision, while of the sons of Shem, again, only one, Aram, is further subdivided. Now it is noticeable that none of these nations particularly singled out are such as have had any close or direct contact with the Hebrews. The identification of Gomer with the Gimirrai who appear in the inscriptions of Assyrian kings being quite certain,’ the subdivisions of Gomer, viz., Ashkenaz,’ Riphath and Togarmah, must likewise represent peoples whose settlements are to be sought in the north- eastern or eastern section of Asia Minor. They belong to the ‘* extreme north,’* have nothing to do with Hebrew history and could only have been of interest to Hebrews because of the general terror inspired throughout the ancient Orient by the threatening 1Cf. Schrader, Cuneiform Inscriptions and the Old Testament, i, p. 62, and Meyer, Geschichte des Alterthums, i, pp. 516 and 543-548. * Distorted from Ashkuza, according to Winckler (Xei/inschriften und das Alte Testament, i, p. 101), who regards them as the Scythians, *Cf, Ezekiel 38, 6. Ashkunaz is only referred to once again in the Old Tes- tament, viz., Jer. 51, 27; Riphath not at all and Togarmah twice in Ezekiel 27,14; 38,6. The parallel Vi/hertafel (1 chr, i, 5-25) dependent on Genesis 10 is, of course, excluded from consideration. 1903.] JASTROW—THE HAMITES AND SEMITES. 177 advance movement of northern hordes during the seventh century B.C. The case is somewhat different with Javan, which is to be identified with Ionia.‘ While none of the subdivisions of Javan enter into direct relations with the Hebrews, with the possible exception of Tarshish,’ until the time of the inpouring of the Greeks into Semitic settlements after the conquest of the Greeks, Cyprus, represented by Kittim, as well as Rhodes, represented by Rodanim,’ must have been at all events familiar names to the Hebrews in pre-exilic days. A certain amount of interest due to commercial relations may also have been attached to the settlements of the Greek archipelago, comprised under the desig- nation, ‘‘ Islands of the Nations.’’ For all that, the sons of Javan have little to do with Hebrew history proper until a comparatively late period. Among the sons of Ham—Cush, Put, Mizraim and Canaan—we might have expected the two last-named to have been taken up in detail and carried down into further subdivisions. If instead of this, it is Cush that is carried down into two subdivis- ions, the conclusion appears justified that in this case, likewise, the point of view is not that of one primarily interested in Hebrew history ; and it is equally remarkable that of the sons of Aram, viz., Uz, Hul, Gether and Mash, the last three are never mentioned again in the Old Testament, while Uz appears only as the home of Job and in a passage in Lamentations (4, 21) where it is used in parallelism with Edom.* To be sure, we have the genealogy of Shem in the line Peleg-Eber once more introduced in P, namely, Genesis 11, 10-26, and this time carried down to the immediate ancestor of the Hebrews, Abram. But the very fact that this is not done in the 1oth chapter is a further proof for the proposition that 1The term is, however, extended to include Greeks in general (see Meyer, Geschichte aes Alterthums, i, p. 492). In a paper read before the American Oriental Society at Washington, April 8, 1904 (to be published in Vol. 25 of the Fournal of the Amer, Or. Soc,), Prof. C. C. Torrey showed that in the book of Daniel (8, 21; 10, 20; 11, 2) and in the first book of Maccabees, as well as in the Talmudic notices, Javan is even used to designate the Greek kingdom of Syria, replacing the earlier usage as, ¢.g., in the 1oth chapter of Genesis, for which we would thus have as a ¢erminus ad quem the fourth century B.C, 2See Haupt’s discussion of the historical and archzological problems connected with Tarshish in his paper published in abstract in the Proceedings of the Thirteenth International Congress of Orientalists (1902), Section v. 3 Jer. 25, 20, is to be excluded, because of the doubtful state of the text. PROC. AMER, PHILOS. SOC. XLIII. 176. L. PRINTED JULY 13, 1904. 178 JASTROW—-THE HAMITES AND SEMITES, [April 4, the enumeration of the thirty-four nations or groups included in P’s Volkertafel is not done from the point of view of one inter- ested in Hebrew history. The situation is just reversed when we come to the other source, known as J, which has been combined by later compilers with P. Though, unfortunately, only a frag- ment of the original Vé/kertafe/ of the Jahwist has been preserved, yet what data there are, including a number of later glosses and other additions, are all of a kind that betray a manifest interest in Hebrew history and hot in general ethnology. EEE. For the introduction to the second version, we must go back to the close of the gth chapter where we read (verses 18-19): ‘*The sons of Noah that went forth of the ark were Shem, Ham and Japheth.' These three were the sons of Noah and of them was the whole earth overspread.”’ The continuation of the genealogical tradition appears chap. 10, 1”: . **to them sons were born after the flood.”’ After which a break occurs and when we again encounter this Jahwistic version (10, 8) we are in the midst of the genealogy of the Hamites, which extends from verses 8 to 19. First Cush, who begets Nimrod, is taken up, then Egypt and finally Canaan with its offshoots. A second break follows and when the Jahwistic source is again resumed, verses 21 and 24-30, the genealogy of Eber the son of Shem is set forth. Fragmentary as the version thus is—the genealogy of Japheth, e¢.g., being entirely wanting—a further analysis points to at least two strata of tradition which, apparently distinct from the Jahwistic Vo/kertafe/, have been combined with it, together with a number of supplementary or explanatory glosses. The little section 8°* to 12, enlarging upon Nimrod and the origin and extension of Babylonia and Assyria, is couched in an entirely different style from 9, 18-19 and from 10, 13-14, and even in this section verse 9", which aims to furnish an explanation for a ' There is added here a gloss,“ and Ham is the father of Canaan,” to pre- pare us for the tale of Ham's disgrace and for the confusion between Ham and Canaan in the curse pronounced by Noah upon his youngest son (vv, 25-27), *The words “and Cush begat Nimrod (8) may belong to the original Jahwistic Valkerta/el. 1903.] JASTROW—THE HAMITES AND SEMITES. 179 popular proverb, is a gloss added to the section itself, and interrupts the context. Again 16-18" represent either a series of glosses or belong to a different source, while the style of verses 21 and 25-30 is so different again from chap. 9, 18-19, etc., that we are forced to assume here likewise a different stratum. Gunkel’ distinguishes these two strata as Jj and Je, on the supposition that the Jahwistic documents in the book of Genesis represent the combination of the original Jahwist with additions from the Elohist. Whether we accept this or not, there can be no doubt that within the Jahwistic version several distinct and originally independent sections are to be distinguished. In accordance with this view, we would have of the original Jahwistic Vé/kertafel only a brief notice about Cush, a fuller one of Mizraim, while in the case of Canaan there is left only the indications of the geographical boundaries of Canaanitish settlements. Still all these three groups are of profound interest to a Hebrew historian, Cush because of Nimrod the representative of the Babylonians and Assyrians, while Egyptians and Canaanites enter of course into Hebrew history at frequent points and at important crises. Taking up the additions to the remains of the original Jahwistic list of nations, it will be found that they fall in the same category of data that have a special interest for the Hebrew historian. The notice about Nimrod specifies the important centres of the Euphrates Valley, Babylon, Erech, Accad (= Agade) and Calneh.’ In agree- ment with the testimony of modern research, the foundation of Assyria is traced back to Babylonia and the extent of Nineveh, ‘‘the great city,’’ with its suburbs is set forth. The introduction of Heth as a son of Canaan (154) may represent already a supplement to the original Jahwistic document, added because of the interest that the Hittites have for Hebrew history,* and to this notice a complete list has been added of the groups of the Canaanitish nations which the Hebrews found upon entering the country and with whom they are thus brought into direct contact.4 Leaving aside variants or further specifications 1 Genesis, p. 77. 2 According to the Babylonian Talmud (Yoma Io*), Nippur. 3 For a summary of the relations between Hebrew and Hittites, see the writer’s article “ Hittites” in the Encyclopedia Biblica, Vcl. Il. 4 It is to be noted, however, that verses 17-18 furnish names of groups out- side of the Hebrew settlements proper in Canaan. 180 JASTROW—THE HAMITES AND SEMITES. [April 4, of the geographical boundaries of Canaanitish settlements, we have lastly the genealogy of Shem, introduced, however, as the opening words show (v. 21), ‘*To Shem also (sons) were born, the father of all the sons of Eber,’’ for the sake of Eber to whom, through the Eber-Peleg line, the Hebrews are directly traced back. Since, however, this genea- logical chain is furnished by P in the following chapter (11, 16-26) the final redactor contented himself in the roth chapter with sup- plying from the J source the genealogical line of the other son of Eber, namely, Joktan. This list of Joktanides (verses 26-30) is most valuable for several reasons. In the first place it furnishes the proof for the thesis that the redactor who combined J and P- uses the former source as a supplement to P and secondly it shows conclusively that J* contained much fuller indications than the other extracts from it used by P might lead us to believe. Indeed the thirteen subdivisions of Joktan represent a much fuller genea- logical chain than any to be found in P which records only seven subdivisions for Japheth and five for Cush.? The special reason why the redactor introduces this long line of Joktanites appears to be because it embodies a varying addition from P which places Havilah among the sons of Cush and Sheba under Raamah the son of Cush (verse 7), whereas the other source includes Havilah and Sheba among the Joktanites (verses 28-29) and thus makes them descendants of Shem. It is hardly reasonable to suppose that so palpable an inconsistency should have escaped the notice of the redactor and it is certainly more plausible to assume that just because of this contradiction between the two sources, both were introduced side by side, in accordance with the general character of historical composition in the ancient Orient which is so largely compilation. The Arabic historians of later days who are the ‘Or FE (i.¢., Jahwist and Elohist) if we follow Gunkel’s view as set forth above (p. 179). 21f it be assumed that the enumeration of the Eber-Peleg line in the rith chapter has been transferred from its proper place in P's Vi/kertafel, it would follow that P’s list may also have originally been somewhat fuller than at present appears, but this would not alter the main proposition that P represents the basis in the toth chapter of Genesis, supplemented by J and possibly other sources, 1908.) JASTROW—THE HAMITES AND SEMITES. 181 natural successors of the Hebrew compilers and redactors would have proceeded in the same way, only they would probably have introduced the second source by the word 4i/a, *‘ others say,’’ and would have summed up the situation by the usual exclamation, ‘¢ Allah knows best ”—which source is correct. How complete the Jahwistic Vé/kertafe/ originally was is, of course, a question to which no definite answer can be given. If the reference to Nimrod as the son of Cush (8*) belonged to the oldest source in J, it would suffice as evidence that at least two branches— Semites and Hamites—were included and this conclusion is con- firmed by the inclusion of Canaan and Mizraim (13-15) but there is no reference in any of the remaining parts of J to Japheth. This may of course have been due to the omission of the Japheth gene- alogy by the redactor who combined J with P, and if this be the case the further conclusion would be justified that J contained nothing of moment with regard to Japheth that was not already mentioned in P. But besides the possibility that J did not con- tain any genealogy of the descendants of Japheth—though in view of the heading Gen. 9, 18 this is unlikely—there remains as an alternative that Japheth may have been included by J under Shem. There are some strong reasons for concluding that such was the case in at least one of the sources worked up by the ‘‘J’’ school of narrators. Attention has long since been directed’ to the cir- cumstance that in the story of Noah’s curse pronounced on his youngest son (9, 25-27) which is attributed by Gunkel* to J°*, the name of the son who is disgraced is Canaan, doomed to be ‘‘a servant of servants unto his brethren,”’ and this is emphasized by a triple repetition of the curse (verses 25, 26, 27), each time with the name of Canaan. It follows accordingly that the three sons according to what is evidently an earlier tradition are Shem,’ Japheth and Canaan. In the poetical fragment of the curse, Shem and Japheth are represented to be in close contact with each other. Accepting with Gunkel,* Gritz’s simple and striking emendation of verse 26°, 1See, ¢eg., Budde, Urgeschichte, p. 300 sg., and the discussion of Gunkel ( Com. to Genesis, pp. 71-76) and Holzinger (Genesis, pp. 91-93), 2 Genesis, p. 71. 8 Or perhaps Eber, See below, pp. 201 and 204. *Z.¢.,p. 78. The change proposed by Gritz merely involves an alteration in the vowels of one word darakh (“blessed”) for which Gratz suggests darekh -.' 182 JASTROW—THE HAMITES AND SEMITES. (April 4, ‘* Bless, O Jahweh, the tents of Shem,”’ we find in the next verse the hope expressed that Japheth ‘‘ may dwell in the tents of Shem.”” Whatever else may be meant by this phrase, it certainly points to a close association of Japheth with Shem. The phrase is intelligible only on the supposition that Shem and Japheth represent two subdivisions of some larger unit in alliance against a common enemy, Canaan; the three—Shem, Japheth and Canaan—so far from representing the nations of the known world, would thus turn out to be originally designations for tribes or clans dividing between them a comparatively restricted strip of territory. Canaan is of course a perfectly definite geogra- phical and ethnic term, and if he is to be the servant of Shem and Japheth, it can only be because he has been or is to be reduced to . servitude and subjection in his own land, and if Shem and Japheth are the subjectors they too must belong to the district in which Canaan lies. Shem in the combination stands for the Hebrews as conquerors of Canaan and whatever may have been meant by Japheth—presumably some allies of the Hebrews—the Japheth in- troduced into the poetical fragment of Noah’s curse is totally different from the Japheth who appears in the 1oth chapter in P as the ancestor of the ‘‘ distant ’’ nations or groups. The later stratum of J no longer knows the subdivisions Shem, Japheth and Canaan. Ham has taken the place of the latter and in order to reconcile the contradiction between the older poem and the later story of Ham’s conduct towards his father, the gloss is added in verse 18 ‘and Ham is the father of Canaan”’ and again in verse 22 the words ‘father of Canaan” after Ham’s name." The story and the poem do not appear to have originally stood in any connection with each other, the latter being here introduced merely as an appropriate climax, just as elsewhere in the Old Testa- ment we find snatches of old poems attached to later narratives “« bless ” (imperative) and the change of ¢/0/@ Shem “God of Shem” to ’Ohode Shem “tents of Shem.” Budde (Urgeschichte, p. 73) proposes to read d°%2akh and to omit ¢/o/2 so that the section would read “ Blessed of Jahweh is Shem,” The objection to this view (though preferred by Holzinger, 4. ¢., p. 90) is the omission of a word whose presence must be accounted for, ' This view seems to me more satisfactory than to regard “ Ham the father” as the gloss which, to be sure, would make Canaan the chief actor as in the original form of the story was the case, 1903.] JASTROW—THE HAMITES AND SEMITES. 183 having no connection originally with the tale itself. Whether the introduction of Ham led to the implied change and to the en- largement of the conceptions connected with Japheth is, again, a question on which it is useless to speculate, and we must content ourselves with the recognition of the wide gulf existing between Shem, Japheth and Canaan on the one hand as they appear in the old poem and Shem, Ham and Japheth as found in the later strata of J and in P. The poem is a fragment of an old composition reflecting tribal dissensions—probably in Palestine—whereas the later figures of Shem, Ham and Japheth belong to the period of an enlarged historical perspective and of more advanced political organization, when, through contact with the nations around, interest was aroused in the larger aspects of humanity as a motley group of peoples and when speculation arose as to the origin of the great variety of nations into which mankind appeared to be divided. This speculation woven in with more or less uncertain traditions and legendary lore finds its first definite expression in a survey of the nations of special interest to the Hebrews and its final outcome in such an elaborately constructed list as is furnished by the present Volkertafel. IV. Coming back, now, to this contrast presented by the remains of an older Vé/kertafel as embodied in J and the later one in P, we are permitted to conclude from the fact that the final redactor of J and P supplemented P’s list by data which bear primarily on those nations with whom the Hebrews came into more or less close con- tact in the course of their history, and since J (with later editions) constituted the source of the compiler for such data, J’s Valkertafel would thus represent the natural intermediate stage between an indifference on the one hand to the determination of the relation- ships existing between the nations of the known world—the feature of the period in which Shem, Japheth and Canaan living in close proximity to one another marked the extent of ethnological inter- est—and the endeavor, on the other hand, to view this relationship 1 £.g , the so-called “Song of the Well’? (Numbers 21, 18) which certainly does not fit in with the narrative in which it has been inserted, and the “ Song of Heshbon ” (74. vv. 28-30) which is a song celebrating the triumph of some peo- ple—hardly the Hebrews—over Moab and which is introduced in connection with a tale of Israel’s victory over Sidon. See Gray’s Commentary to Numéers, PP- 301, 302. 184 JASTROW—THE HAMITES AND SEMITES. [ April 4, from the broadest standpoint possible to an ancient writer or to a school of ancient writers with imperfect ethnological conceptioris and still swayed to a certain extent by various subjective factors. If, therefore, Japheth formed part of J’s Vélkertafel, we may feel reasonably certain that it did not concern itself with such nations as Gomer, Gog, Ashkenaz, Riphath and Togarmah, to mention only some groups with whom Hebrew history has nothing to do, but at the most with such groups as Tarshish, Cyprus and Tubal and Meshech, with whom at a certain period the Hebrews had at least commercial relations. Leaving this question aside as impossible of more definite determination, the remarkably inclusive though compact character of P’s list, drawn up from a point of view which betrays no special interest in Hebrew history, suggests a for- - eign source for the list itself, or at all events points to foreign influ- ences at work in its composition. The Priestly Code, being an exilic production, of which at least the substantial elements were drawn up in Babylonia, it would be natural to seek in it influences due to the Babylonian environment. The earlier political relations of their own people with Egypt and Assyria would be sufficient, with the rise of the historical sense, to arouse in the minds of Hebrew writers an interest in nations lying outside of their own immediate circle, but this interést would be materially strengthened under such conditions as confronted the Hebrew exiles settled in the Euphrates Valley. With the national catastrophe putting an end for the time being to their own political history, the Hebrews were in a peculiarly favorable position for realizing what the world meant to a world-power such as Babylonia, which had undertaken to still further develop the legacy of con- quest and subjugation bequeathed to her by her rival Assyria, had become in the sixth century. They found themselves in a country which stood for the ideal of world conquest, and which had taken decisive steps for many centuries toward the realization of this ideal. The Assyrians and the Babylonians had come into direct contact with distant nations to the north, south, east and west, and, although their relationship to those nations had generally been hostile, they had, yet, by the encouragement of international com- merce brought about a closer affiliation between the peoples of the ancient world, than is ordinarily recognized. It would have been strange indeed if, under such surroundings, the Hebrews had not been led to modify and enlarge their views of the complicated 1903.] JASTROW—THE HAMITES AND SEMITES. 185 constitution of mankind. The inscriptions of the Assyrian kings abound in geographical details," and the interest of both Babylonians and Assyrians is still further attested by the numerous geographical lists? that have been found in Ashurbanapal’s Library and elsewhere. While it is true that these lists, as a rule, were prepared for practical purposes, in connection with the campaigns or as tribute lists or as exercises to serve in the training of scribes, yet a theoretical interest must also in the course of time have been awakened and some of the lists clearly betray such interest. What applies to Assyria is true also of Babylonia, with perhaps this difference: that in a land like the latter in which culture had reached a higher level than in the north, the theoreti- cal or, as we might also put it, the scientific interest must, if any- thing, have been much stronger. That the intellectual class among the Hebrew writers was acquainted with Babylonian literature admits scarcely of doubt,® and whether the compilers of the Priestly Code actually had some cuneiform models before them to serve as the bases for such a list as is found in P, it is certainly permissible and indeed a most reasonable supposition to attribute to Babylo- nian-Assyrian influence the striking feature of P’s list that it deals so largely with groups of peoples that are of interest to Babylonian- Assyrian history and of scarcely any at all to Hebrew history. So of the sons of Japheth, Gomer, Madai, Tubal and Meshech occur more or less frequently in Assyrian inscriptions,‘ and to these we may add Ashkenaz,° and perhaps Togarmah.* Nor can it be entirely accidental that so many of the groups included under Jap- heth should be encountered again in the exilic prophet Ezekiel living in Babylonia. He refers to Gomer, Gog, Javan, Tubal, 1A glance at the Indices to such works as Schrader’s Xeilinschriften und Geschichtsforschung, Delitzsch’s Wo Lag das Paradies, and Winckler-Zim- mern’s Ketlinschriften und das Alte Testament will suffice to show how large the geographical horizon represented by the cuneiform annals is, 2 £.g., Rawlinson, ii, 50-53. $See, ¢.g., D. H. Miiller’s Zzechiel Studien, pp. 56-62, who gives some inter- esting illustrations that seem to point conclusively to Ezekiel’s acquaintance with Babylonian literature. See also Winckler’s paper, ‘* Der Gebrauch der Keil- schrift bei den Juden” (A/torientalische Forschungen, iii, 1, pp. 165-174). *See the Indexes in the works above referred to. 5See above, p. 176, note 2, and also Baer’s Lisri Danielis, Ezrae et Nehe- miae (Leipzig, 1882), p. ix. 8See Delitzsch’s Wo Lag das Paradies, p. 246, and Jeremias’ A. 7. im Lichte des Alten Orients, p. 152. 186 JASTROW—THE HAMITES AND SEMITES. [April 4, Meshech and Togarmah,’ and in general this prophet is distin- guished by the very wide range of his geographical knowledge. We are, therefore, justified in concluding that Babylonian influence and contact with the intellectual atmosphere of Babylonia are responsible for the display of geographical interest and learning in P’s Vélkertafel and in Ezekiel. On the other hand, for the knowledge of Ionia (Javan) it was not necessary to turn to Babylonia or Assyria, for as already suggested,? commercial relations between Palestine and the islands and districts lying to the west up to distant Tarshish would account for the knowledge of the chief settlements in the Greek archipelago and regions beyond. This knowledge may well have existed among the Hebrews in pre-exilic days, and the view here maintained by no means implies that all the geographical learning displayed in P comes from contact with Babylonia, but merely that, apart from certain direct influences, the enlargement of the ethnological horizon of Hebrew writers and the impulse to draw up such a Volkertafel as is found in P can best be accounted for by the new factor that entered into the intellectual life of the Hebrews through their settling in the Euphrates Valley. Be this as it may, the political contact of the Hebrews with those groups enumerated as sons of Javan did not begin until the period of Greek conquests in the Orient, and, unless we choose to bring the compi- lation of P’s Vélkertafel down beyond the age of Alexander the Great, which on other grounds is improbable, we are forced to con- clude that all the nations enumerated under Japheth are to be placed in the category of peoples with whom the Hebrews up to the time of the composition of the Priestly Code had practically nothing to do. The division of Japheth into two branches, (a) Gomer and offshoots and (b) Javan and offshoots, merely represent from the point of view here maintained the distant nations dwelling to the northeast and north on the one hand, and the groups to the west and northwest on the other, more particularly the inhabitants of the Grecian islands, and those settled along the coast of Asia Minor. V. Coming to the Hamitic genealogy, the wavering of traditions "See, ¢y., chapter 38 of Ezekiel, See above, p. 177. 1903.] JASTROW—-THE HAMITES AND SEMITES. 187 in the case of the sons of Cush makes it difficult to reach any definite conclusion as to the point of view which guided the com- piler of P’s Vélkertafel. Besides the contradictions already pointed out in the case of Havilah and Sheba,' it isto be noted that Dedan, who in P appears in the genealogy of Cush, is, according to Genesis 25, 3, included with Sheba in the genealogy of Abraham. That in the mind of P, the sons of Cush represent certain nations of south- ern and central Arabia, with perhaps an inclusion of some groups lying along the eastern coast, is about all that can be said with any degree of definiteness.?. That Put in P’s list represents primarily the western coast of Africa, from upper Egypt and southwards to Somali (though also applied to the corresponding Arabian coast land), has now been definitely shown.* We would thus obtain a point of union for Cush and Put in the circumstance that they rep- resent remote people in the mind of P, lying to the extreme south. This might be extended to Mizraim, but certainly Canaan, which has always been the stumbling block in attempts at recognizing any system in the grouping of Hamites, cannot be placed among the nations of the south without our having recourse to the most 1See above, p. 180. 2See Jeremias, 4. 7: im Lichte d. Alten Orients, p. 155, and Glaser, Skizse der Geschichte und Geographie Arabiens, ii, pp. 387-404. It is unnecessary to pass over to the African coast for the identification of any of the seven groups, though it is certain that P as well as J, in accord with the general usage of the Old Testa- ment, regards Cush also as a designation of Nubia. The term seems to be some- what indefinitely used for the extreme’south (or what appeared to be such to Hebrew writers) without a sharp differentiation between southern Arabia and the corre- sponding district on the African coast. On Cush as a designation of a part of Arabia in the Old Testament, and in the Cuneiform Inscriptions, see Winckler, Keilinschriften und das alte Testament,p. 144-145, summarizing views expressed in his essay, “ Musri-Meluhha-Ma’in,” i and ii (M/t¢thetlungen d. Vorderasia- tischen Geselischa/t, Berlin, 1898, pp. 47 sg.), and the same author’s A/¢testament- liche Untersuchungen, p. 165. This double nomenclature of Cush may well be supposed to rest on traditions of an ultimate close relationship between the settlements in Africa and those of southern (and extending into central) Arabia ; and if there is any value to be attached to the precise form given to the tradi- tion in the Old Testament, the conclusion might be drawn that the “ Arabic ” settlements represent the offshoot, 7.¢., “sons ’’ of the African Cush—a view that on the whole seems more plausible than the contrary hypothesis. 8See W. M. Miiller’s Asien und Europa, chap. vii. That it designated pri- marily Arabia is the view of Meyer, Gechichte des Altertums, i, p. 86, while Glaser, Skizze, etc., ii, pp. 405, 406, proposes southern Arabia and the east coast of Africa. 188 JASTROW—THE HAMITES AND SEMITES. [April 4, risky and hazardous conjectures.’ It is to be noted that Canaan occupies the fourth and last place among the sons of Ham, which of itself raises the suspicion that its addition is due to an after thought, and that, moreover, P does not follow up the genealogy of either Mizraim or Canaan, so that the later redactor was obliged to supply the omission from J. I venture to suggest, there- fore, that we have in the addition of Canaan the first betrayal of the compiler’s subjective point of view. Under the influence of the same hostile spirit toward the Canaanites which manifests itself in the old poem embodied in J, but with the extension of this hostility to a larger group—to Ham, which was substituted for Canaan—the compiler of P’s list places Canaan in the group now associated with the accursed nations, but which was originally intended merely to represent the remote nations of the south, as Japheth rep- resented remote nations of the west, north and northeast. That even a learned compiler living in Babylonia, and actuated primarily by a scholastic aim to draw up an elaborate scheme of a series of nations and peoples in illustration of his theory that all mankind can be traced back to a single ancestor, should be subject to the deeply imbedded hostility existing from the days of the Hebrew invasion of Palestine between Hebrews and Canaanites is surely not surprising. Such a limitation of the mental horizon is precisely of the kind that we would have aright to expect. Removing Canaan from the group, we would have the Hamites consistently represent- ing the remote nations of the south, as the Japhethites represent the remote nations of the west, north and northeast. VI. Leaving aside for the moment the problem involved in the change of sentiment which converted the Hamites into a group synonymous with the ‘‘ accursed”’ nations and turning to the gen- ealogy of Shem, it is noticeable that the beginning is made with Elam, lying immediately to the east of Babylonia, and that the group is closed with Aram, which appears to be a general designation for the district lying to the west and northwest of Babylonia and Assyria. We now know that the political relations between Elam ‘As, ¢g., the view maintained by Dillmann (Genesis, p, 179) that the inclu- sion of Canaan among the sons of Ham rests upon the knowledge that the Canaanites came from a southern district, 1908. JASTROW—THE HAMITES AND SEMITES. 189 and Babylonia date back to a very early period,’ and that in fact the history of the one district is so closely entwined with the fortunes of the other that it would be quite as natural to group Elam and Babylonia together as to place Babylonia and Assyria side by side. Linguistic and ethnic differences between Elamites and Babylonians would not obtrude themselves to the mind of an ancient writer in the face of such close political associations as bound Elam and Babylonia together. Again, a grouping which begins with Elam as the eastern out- lying province of Babylonia and ending with Aram as the western limit would be intelligible from the standpoint of one living within the district of Babylonia, and this view is confirmed by the intro- duction of Asshur immediately after Elam. Moreover under Aram, subdivisions are recorded—Uz, Hul, Gether, Mash*—that play no part whatsoever in Hebrew history, and could have been of interest only to Babylonians and Assyrians as representing districts lying beyond the Euphrates, and with which their armies would come into contact in the course of expeditions to the west or by which they might at one time or the other have been menaced. At all events, Aram designates a miscellaneous group of peoples whose settlements form the western boundary of Babylonia and Assyria proper, ard so far we would have as the point of union in the enumeration of the sons of Shem, the settlements in the immediate environment of Babylonia and Assyria—to the east and west respectively. This view is not contradicted by the mention of Arpachshad immediately after Asshur, for however we wish to account for this name, the last element 4-sh-d is certainly in some way connected with Xashdim—the designation of the Chaldeans. Of the various explanations offered,* the most plausible is to divide the word into two elements, a-r-6 which may be identified with Arrapachitis (= Aréakha) and k-sh-d which is Chaldza, so that we would have two distinct districts that have by an error been ‘See De Morgan, “ L’Histoire de Elam ” (Revue Archeol., 3em. Serie, Vol. 40, pp. 149-171). * For proposed identifications of Uz, Hul and Mash see Gunkel (Gemesis, p. 142), Holzinger (Genesis, p. 105) and Glaser (Ské2ze, efc., pp. 411-422). The situation of Gether is entirely unknown. 8See Holzinger, Genesis, p, 105, and Gunkel, /.c., p. 143, who accept Cheyne’s view of the division of the word (Zeits. f. Alttest. Wiss., 1897, p. 190), into Arpach and Keshed. 190 JASTROW—THE HAMITES AND SEMITES. [April 4, merged into one. However this may be, so much is certain that Arpachshad is still to be sought within or near the district repre- sented by Babylonia and Assyria. More puzzling than Arpachshad is the fourth subdivision Lud, and no entirely satisfactory explanation of its occurrence here has as yet been offered. Certainly Lud in P’s list can have nothing but the name in common with the Lud that occurs in Isaiah 66, 19, and in Ezekiel 27, ro and 30, 5—which is clearly Lydia in Asia Minor—and unless we assume (as I am in- clined to do) that the introduction of Lud in Genesis 10, 22 is due to an error— Arpachshad w*-Lud being here (verse 22) superinduced by Arpachshad ya-lad in verse 24’— we must provisionally accept the possibility of there having been a district Lud between the Babylonian-Assyrian dis- trict and what P understood by Aram. For the present, there are no substantial reasons for questioning on this account the thesis here maintained that in P the Shemites represent Babylonia and Assyria and the groups adjacent, in contrast to the Japhethites and Hamites who represent the remote nations in the various directions of the compass. We may, therefore, conclude that P’s list, taken as a whole and leaving aside more or less obscure details which do not, however, upset the general conclusion, betrays the learned compiler whose geographical horizon has been enlarged by becom- ing subject to his Babylonian environment. In addition to gather- ing some of his geographical knowledge from Babylonian docu- ments or through intercourse with the learned scribes of Babylonia, his general point of view in his grouping of nations has regard for interests affecting Babylonia and Assyria, as in the case of the northern and northeastern branches of the Japhethites, or is deter- mined, as in his grouping of Shemites, by his residence in Babylonia, The purely scholastic character of the list is interfered with only by the addition of Canaan to the Hamitic group, the introduction Wiedemann supposes (Geschichte Aegyptens, p. 24), that Lud is the original form of Rut which with a « denominative ” ending—i,e,, Ruten—occurs in Egyptian inscriptions as the designation of Syria and Palestine, See however the objections to this conjecture in Schrader’s Cuneiform J/nscriptions and the Old Testament,i,p.99. Nor is Jensen's proposition to read Lubdi (adopted by Jeremias, A. 7, im Lichte des alten Orients, p, 170), at all satisfactory, 1903 J JASTROW—THE HAMITES AND SEMITES. 193. of which is due to the hostility existing between Hebrews and Canaanites.! Taking up now this departure on the part of the compiler of P from the scholastic principles that guided him in drawing up the list, it is clear that he could only have been led to destroy the harmony of his scheme by placing Canaan under Ham instead of, according to their proper position, next to Aram, if the view had become general that the Hamites represent the ‘‘ accursed ”’ nations. VII. To justify this assumption, which involves a radical change from the original conception of the Hamites as the rations of the remote south, it is necessary to find other evidence for it. Such evidence is forthcoming not merely in the narrative which substitutes Ham for Canaan, but also in J’s grouping so far as his Vé/kertafel has been preserved. We have seen that J enlarges the curse originally pronounced upon Canaan into a general denunciation of a larger group whom he calls Hamites. At the same time, he does not venture to alter the ancient tradition entirely but makes a compromise by including Canaan under Ham. Whatever the source may have been whence J derived the name of Ham, for him this youngest on of Noah has clearly come to be synonymous with those nations which are par- ticularly obnoxious to him. Let ussee whom J places in this group. We have in the first place Nimrod whom he connects with Cush as against P who does -not mention Nimrod, but who places seven other nations, representing groups settled in Arabia, among the sons of Cush.*? Nimrod, however, as verses 10 and 11 clearly show, is in J’s list the representative of Babylonia and Assyria—nay the founder of these empires, in marked contradistinction therefore to P who, as 1It is only proper to note that the view which assumes Canaan’s place among the Hamites to be due to feelings of natural hostility was maintained by older: writers as, ¢.g., Sprenger (Geographie Aradbiens, p. 294 seg.) who lays strong emphasis on the point, but since the days of Dillmann has been generally aban- doned. The attempts, however, that have been made to account for the place assigned to Canaan are singularly inadequate. Recent writers either ignore the point entirely or content themselves, like Holzinger (Genesis, p. 96), with the suggestion that the inclusion of Canaan among Hamites is merely characteristic of the prevailing ignorance among the Hebrews in matters pertaining to ethnology. 2 See above, p. 187. \ > 192 JASTROW—THE HAMITES AND SEMITES. [April 4, we have seen, places these nations among the sons of Shem. If Nimrod is a Hamite, it follows that Babylonians and Assyrians are Hamites and the attitude towards Nimrod implied in thus placing him among the Hamites is clearly indicated by the gloss (verse 9) ‘‘he was a mighty hunter before Jahweh” where the words ‘‘ before Jahweh’’ indicate as is now generally recognized by commentators’ ‘‘ in defiance of Jahweh,”’ implying an opposition of some kind to Jahweh or if that is going too far, as, at allevents, carrying on a pursuit which was not pleasing in the eyes of Jahweh. Whatever the original force of the phrase ‘‘ mighty hunter’’—concealing perhaps some reference to an ancient myth— may have been, to the one who introduced the gloss in J’s list of nations Nimrod was a conqueror, a ‘‘hunter’’ of spoil, as it were, fired by the ambition to extend his dominion. As a conqueror he, therefore, appears in the following verses where the enlargement of his kingdom is referred to and the extent of Babylonia and Assyria is indicated by the mention of the chief cities of both districts. To J, therefore, the chief if not the only interest attaching to Cush lies in his being the ancestor through Nimrod of Babylonia and As- syria and whatever other nations—if any—were included by him under Cush. His motive for making Babylonia and Assyria descend- ants of Cush was not geographical position, nor is it at all likely that he had in mind a district by the name of Cush to the east of Babylonia whence in his opinion Babylonians and Assyrians came? —though it may be admitted that the notice rests ultimately on a confusion between two Cushs*—but he was actuated solely by the desire to place Babylonians and Assyrians among the Hamites 1 See, ¢.g., Budde, Urgeschichte, p. 393; Holzinger, Genesis, p.99. Renan, too, explained the phrase as indicating opposition to Jahweh. Compare also the phrase ‘‘a great city to God” (Jonah, 3 3), equivalent to a * godless city,” *So ¢g., Winckler, Alttestamentliche Untersuchungen, p. 149. Cf. Gunkel, Genesis, pp. 81-82. * For our purposes it is immaterial whether Cush in the mind of the writer who added the section about Nimrod meant the African or the Arabic Cush; and even though some faint tradition of a third « Babylonian ” Cush (7.¢., the Cassites) underlies the tale, it is certain that the writer has the same Cush in mind as in P (verse 7). Delitzsch's view (Wo Lag das Paradies, p. 52 seq., and pp. 127- 129) of a close historical connection between the “ Babylonian” and « African ” Cush is untenable, though he correctly places the seven subdivisions of Cush in Arabia and not in Africa, See above, p. 187. 1903.] JASTROW—THE HAMITES AND SEMITES, 193 under his general view that the latter represent the accursed nations. If we turn to Hebrew history, we will find in the relations exist- ing between the Hebrews and the Babylonians and Assyrians the all-sufficient motive for this hostility. Babylonia which exercised a control over Syria and Palestine at a very early date as the Tel El] Amarna tablets show,’ until she was obliged to yield to Egypt and to concentrate her efforts on the endeavor to check the growing power of Assyria, must have been regarded as a natural menace to the Canaanitish settlers in Palestine even before the Hebrews entered the land. The latter therefore inherited from their predecessors a feeling of hostility towards Babylonia and not differentiating Baby- lonia sharply from Assyria, the bitter feeling towards both would be accentuated by the subsequent course of events.” From the ninth century on, the two Hebrew kingdoms were exposed to fre- quent attacks from the military expeditions undertaken by the Assyrian conquerors. The fall of the northern kingdom in 722 B.C. and the practical subjection of the southern by Sennacherib and his successors further strengthened this hostility, which found a forcible expression in the ‘utterances of the pre-exilic prophets and is reflected in the grouping of Babylonia and Assyria with the **accursed’’ nations in J. It is not necessary for our purposes to assume that the form given to the feelings of resentment against Babylonia and Assyria actually presupposes the destruction of the southern kingdom, for long before this catastrophe the feelings must have been sufficiently strong to prompt a writer to regard Babylonians and Assyrians as ‘‘ accursed ’’’ in the eyes of Jahweh, so that the little section inserted in J verses 8-12 may be, like J’s list, of pre-exilic date; but we may well suppose the post-exilic redactor who combined J with P to have been still further incensed at the recollection of the havoc wrought by Assyria and Babylonia, the one in bringing about the downfall of the northern kingdom and the other the extinction of political life in the south, to prompt him to preserve from J—in its final form—the notice which groups 1See Winckler, Ket/inschriften und das alte Testament, pp. 23-25 and 192 seq. *See for the general relationship between Babylonia and Assyria, and the two Hebrew kingdoms, Winckler, Kez/inschriften und das alte Testament, pp. 258- 280 ; also for the cuneiform texts bearing on the subject Schrader’s Cuneiform Lnscriptions and the O, T., i, 176-ii, 59, and Winckler, Keilinschriftliches Textbuch 2um alten Testament (2d ed., Leipzig, 1903), pp. 14-55. PROC. AMER. PHILOS, SOC. XLIII. 176. M. PRINTED JULY 13, 1904. 194 JASTROW—THE HAMITES AND SEMITES, [April 4, these two peoples among those whom Jahweh himself has cursed— much in the same spirit that leads to the retention (Genesis 19, 30- 38) of the scandalous story of the origin of Moab and Ammon— two other bitter enemies of Israel—from an incestuous union of Lot with his daughters, as a bit of tribal satire, calculated to expose these peoples to the humiliation and contempt of their rivals.’ After Nimrod, we find Egypt introduced in J. Among the Hamites we have seen that the grouping of Egypt with Cush and Put in P fits in with the latter’s general view that the Hamites represent the nations of the remote south, but J for whom Cush is neither southern Arabia nor Nubia does not appear to have had such a scheme in mind, and it is in keeping with the spirit of the narrative at the close of the 9th chapter that J’s motive in add- ing Egypt to abylonia and Assyria-among the Hamites was again the desire to illustrate the truth and the justification of the view that the sons of Ham are the ‘‘accursed”’ nations. It is only necessary to mention the name of Egypt in order to conjure up the picture of the hostility towards it that crops out in every section of the Old Testament. The recollection.of Egyptian oppression is so strong in the Old Testament as to become almost a part of the Hebrew religion. An old nomadic sheep-offering festival com- bined with an agricultural spring festival, the latter adopted by the Hebrews from the Canaanites, becomes associated in the Pen- teteuchal codes* with the deliverance from the hated yoke of Egypt. The Decalogue begins in both versions that we possess with the description of Jahweh as the god who brought his people out of Egypt (Ex. 20, 2; Deut. 5, 6), and according to the Deuteronomistic version or recension of the Decalogue, the most characteristic institution of Judaism—the Sabbath as a day of rest 'So sccording to the best of the modern commentators (see Holzinger, Genesis, p. 158). Somewhat different is Gunkel’s view (Genesis, pp. 197-198), who believes that the story was originally told as an illustration of the favor and grace of the Deity in saving Lot as the ancestor of Moab and Ammon from the general destruction and in providing for this unusual method of securing offspring. Granting this, it is still evident that in the mind of the Hebrew writers, the story assumes a lowering ani contemptuous aspect—to be compared with the bitter taunts and satires to be found in ancient Arabic poems when they deal with tribal hostilities. See ¢., Goldziher, Muhammedanische Studien (Halle, 1888), 1, pp. 43-50. *See Baentsch, Com, fo Exodus, pp, 88-91; Robertson Smith, Religion of the Semites, pp. 445 seg. 1903.] JASTROW—THE HAMITES AND SEMITES. 195 from all labor—is instituted to serye as a reminder to the people of the conditions under which they lived in Egypt (Deut. 5, 15). If we turn to the Prophets, we find Egypt invariably associated with cruelty, deceit and oppression.* Pharaoh becomes a type of the persecutor and of the oppressor. Egypt is therefore placed like Babylonia and Assyria in the same category as Canaan—with the ‘*accursed’’ races. It so happens, as already pointed out, that the position of Egypt accords with the geographical scheme that P adopts for the Hamitic nations; and while, in view of this, we are not justified in attributing to this compiler a motive of national hatred in placing Egypt with Cush, J, who does not appear to have had such a geographical system and for whom Ham is merely the larger term for Canaan which permits him to place under one category a whole series of nations who were hostile to his people, and who in his opinion are responsible for the dark pages in _pre-exilic Hebrew history, is evidently actuated by such motives of national hatred in associating Egypt with Canaan ; and as already intimated, the compiler who combined J with P, likewise, no longer occupies the objective and more purely scholastic standpoint of P, and takes over therefore from J the extended notices about Egypt and Canaan in order to point out in detail all those who belong to the ‘‘ ac- cursed ’’ sons of Ham. : VIII. This spirit of hostility crops out again in the inclusion of the Capthorites (verse 14) where the addition of the gloss ‘*‘ whence came the Philistines’’ reveals the animus of the compiler. Cap- thor, as Prof. W. Max Miiller* bas shown, is a term of indefinite character but which certainly included Cilicia and adjacent parts of the Asia Minor coast, and even a writer of so limited a range of ethnological and geographical knowledge as J, granting that he no longer knew the exact distinction of Capthor,* could hardly have supposed the Capthorites to belong in the same category with the 1 It is sufficient to refer to such passages as Isaiah 11, 15, and chap. 19; Eze- kiel, chap. 30; Jeremiah, chap. 46; Amos 8, 9. 2 Asien und Europa, p. 347, supplemented by the same writer’s Studien zur vorderasiatischen Geshichte, ii (Mitteilungen der vorderasiatischen Gesell- schaft, 1900), pp. 6-11. 8 That in accord with prevailing views or traditions he identified Capthor with Crete is, on the whole, more than likely. 196 JASTROW—THE HAMITES AND SEMITES, [April 4, subdivisions of Egypt, whose mention precedes that of Capthor.’ Moreover, the position of the Capthorites at the close of verse 14” suggests (as we have seen to be the case in other instances of nations placed at the end of a series of groups), a later addition to what precedes, and the gloss indicating the origin of the Philistines in accord with the tradition recorded in Amos? and Jeremiah, and which is also found in Deuteronomy,‘ unmistakably reveals the purpose of the addition. Next tothe Canaanites, whom the Hebrews had to drive out before they could acquire a foothold in Palestine, the Philistines constituted the most serious obstacle to the growth of an independent Hebrew state. Prior to the days of Saul, we have three distinct periods of Philistine aggressiveness with disastrous results to the Hebrews (Judges chapter 10; Judges chapter 13; 1 Samuel chap. 4). Hostilities continued with changing fortunes through the days of Saul and David. Solomon appears to have held them in check, but after his death they regained their independence and continued to be a source of annoyance to Israel if no longer a serious menace. The Capthorites, accordingly, as identical in the mind of the one who added the gloss with the Philistines are ranked like Canaan, Babylonia, Assyria and Egypt with the ‘‘ accursed ”’ nations, who were assigned this character because of the bitter feel- ings of hostility of the Hebrews towards them. The ‘ accursed ’’ nations thus turn out to be the enemies of the people of Jahweh, whose opposition is looked upon as a defiance of Jahweh himself. Outside of the addition of Capthorim in verse 14, the subdivis- ions of Egypt, enumerated in the verse in question, obscure as some of the names are, are introduced as an exhibition of learning from purely scholastic motives, which J is also willing to display where they do not interfere with his nationalistic likes and dislikes. On the other hand, it is in all probabilities a personal interest that is displayed in the enumeration of the clans constituting the sub- divisions of the Canaanites. This enumeration is not set forth in the form of a genealogical chain and the proof that the list itself 'Verses 13-14%. Of the six subdivisions of Egypt, only two, Lehabim = Lybians, and Pathrusim <= Upper Egypt, are certain, but that the other four, all probably more or less corrupt forms, represent sections or nomes of Egypt is gen- erally admitted, For further attempts at identifications see Holzinger, Genesis, pp. tol-102, * Amos 9, 7. 5 Jeremiah 47, 4. * Deut. 2, 23. 1903.] JASTROW—THE HAMITES AND SEMITES. 197 represents a later gloss, incorporated however with J and not belong- ing to P, is furnished by the gentilic form (Jebusite, Amorite, etc.) given to the nine Canaantish subdivisions." The subdivisions _ themselves further emphasize and illustrate the point of view from which the Canaanites are regarded in J. Of the nine subdivisions, four (Jebusite, Amorite, Girgasite, Hivite) belong to the seven nations of Palestine, with whom marriage is forbidden in the Pen- tateuchal codes,? and with whom no alliance of any kind is to be made; and since it is likely that the Hamathites, referred to in Gen. 10, 18, stand for the Hittites of Deuteronomy 7, 2, we would have five of the ordinary seven subdivisions of Canaanites enume- rated in this addition to J. The author of this addition, well acquainted with the various Canaanitish settlements in Palestine, introduces these five because of his special interest in that part of Palestine with which Hebrew history is especially concerned, and ' which was promised to them by Jahweh as their future possession (cf. Gen. 15, 18-20). In adding the Arkites, Sinites, Arvadites and Zemarites, which play a less conspicuous part in Hebrew history, he reveals his learning and scholastic interest, whereas on the other hand verse 15, which reads ‘* And Canaan begat Sidon his first born and Heth,”’ reveals the original force attached to Canaan as embracing the Pheenicians as well as those settled in the interior. The style of this verse shows that it belongs to the original J document, though there are reasons for believing that the verse has not been preserved in its original form. If Sidon is mentioned as ‘ the first born’’ we would expect other sons to have been included in the genealogy ; and, again, the words ‘‘ and Heth ”’ impress one as a iater addition of the same nature as the additions at the end of verse 14° and elsewhere. The suspicion is, therefore, raised that ‘‘ Heth’’ has been attached to Canaan from the same motive of nationalistic senti- 1Jebusite, Amorite, Girgasite, Hivite, Arkite, Sinite, Arvadite, Zemarite, Hamathite, The traditions in regard to the forms of these names seem to be pretty definitely established, except in the case of Sinite, for which the Greek version has //asennite and the Aramaic (Targum Onkelos, ed. Berliner, Berlin, 1884) Antusite. *See ag., Ex. 34, 11-16; Deut. 7, 1-3. Cf. also Gen. 28, 1-S—a narrative that well reflects the bitterness of the feeling toward the Canaanites. 3See above, pp. 179, 188, 196. 198 JASTROW—THE HAMITES AND SEMITES. [April 4, ment which relegates all enemies of Israel who hindered the advance of the latter among the ‘‘accursed’”’ nations. It is needless for our purposes to enter upon the vexed question whether the B*ne Heth, settled in southern Palestine, are to be identified with the Hittites in the northeast, where Hamath formed one of the centres of their settlements." The Hebrew writers, as is quite evident; considered them identical, and although those in the south enter into friendly relations with the early Hebrew invaders, as illustra- ted by the traditions regarding Abraham’s dealings with the B*ne Heth,’ those in the north are included among the enemies with whom no alliances of any kind are tobe made. The term “ Heth”’ may indeed have been introduced by the one who added it to Sidon to include the entire interior of Palestine, which a later glossator not satisfied with so vague an expression amplified by the specifi- cation of the nine subdivisions in verse 16-18. However this may be, the addition of Heth and the further specification of nine subdivisions, whether originally intended as specifications of Hittites or of Canaanites, are prompted and retained by the desire © to make it perfectly clear that the groups with which the Hebrews were to make no entangling alliances of any kind, whether social or political, belong to the ‘‘ accursed ’’ Hamites. This same motive is further illustrated in the indication of the boundaries of the Canaantish settlements. ‘Taking in the Phoeni- cian coast to Gerar, or according to the variant to Gaza,‘ he car- ries the eastern boundary to Sodom and Gomorrah. I venture to suggest that this specification is not prompted by purely scholastic interests, but from a desire to leave no doubt, on the one hand, as to the inclusion of the hated Philistines, represented by Gerar and Gaza, among the Hamites, and, on the other, to point out by the 'That the term “ Hittites’ was used to embrace large groups of peoples that entered Syria and Palestine from the mountain districts of the north and north- west is now generally recognized, The vagueness of the nomenclature complicates the historical and ethnological problems, but it may be said that what evidence is available does not militate against regarding the northern and southern Hit- tites of Palestine and Syria as belonging to the same general group. * Genesis, chap, 23. * See above, p. 197. *Verse 19. It matters little whether we take Gerar or Gaza as the gloss, though the former, about six miles farther south of Gaza, being less well known, probably represents the original reading, to which a glossator added as a memo- randum * Gaza,” as a better known boundary of Philistine settlements, 1903. JASTROW—THE HAMITES AND SEMITES. 199 introduction of Sodom and Gomorrah that the inhabitants of this district as well as two other peoples particularly obnoxious to the Hebrews, namely, Moab and Ammon, whose origin, according to the libellous tradition recorded in Genesis 19, 30-38, is dis- tinctly connected with Sodom and Gomorrah, also belong to the ‘‘accursed’’ nations. This tale follows immediately upon the story of the destruction of Sodom and Gomorrah, evidently with the intention of associating Moabites and Ammonites,—whose hostile relations to the Hebrews are illustrated in many a page of the Old Testament,—with wickedness and shameful immorality. However this may be, Sodom and Gomorrah are for J, as for the Hebrew _ prophets, the type of all that is ‘‘ accursed,’’* and for this reason those who dwelt in this region are singled out as belonging with peculiar appropriateness to the sons of Ham. Naturally, there are innumerable details in the early history of the Hebrews, as also in the later periods, which escape us so that it is no longer possible to determine the full extent to which this motive of national dislike influenced the school of writers, the result of whose work is to be seen in J’s list as modified by later additions, insertions and glosses, but enough has been shown to justify the proposition that in contradistinction to P, who betrays not only a much broader geographical and ethnographic knowledge but also greater objectivity, J and the school that he represents are largely, if not completely, under the spell of the character given to Ham at the close of the 9th chapter of Genesis. For J, Ham is not an ethnic unit nor a designation for a group of peoples settled in a certain section of the known world, but a kind of ethnological purgatory to which all those nations are assigned,—Babylonians, Assyrians, Egyptians, Canaanites, Philistines, Sodomites, Gomor- rites,—who have merited this fate in the mind of the writer by their hostility to Jahweh’s people, and as the cause of the misfortunes, hardships, struggles and catastrophes in the career of the Hebrews. On a supposition of this kind we can account for the jumble of such heterogeneous groups as Canaanites in Palestine, Egyptians in the South, Babylonians and Assyrians in the East, and Philistines in the West into one category, unless, indeed, we are prepared to commit exegetical suicide by assuming that no principle whatso- 'Cf., ¢.g., Isaiah 1,10; 3,9; 13, Ig. Jer. 23,14. Amos, 4, 11, and Zeph. 2, 9, where Moab and Ammon are compared to Sodom and Gomorrah. 200 JASTROW—THE HAMITES AND SEMITES. [April 4, ever presided over the grouping. The view here advanced of the different conceptions regarding Ham by the J-group of writers from ‘that which is found in P also accounts in a rational and, I believe, in a satisfactory way for the manifest contradictions between J and P as, ¢.g., the grouping of Asshur with Cush by the former and with Shem by the latter. IX. In conclusion, a few words about the genealogy of Shem in J and P, which will further illustrate the thesis here main- tained. If Ham in the mind of the nationalistic J is the type of the ‘‘accursed’’ son, Shem appears with equal distinctness as the favored one. This view is clearly brought out in the blessing over Shem (Gen. 9, 25-27). The double mention of Shem already shows this, and whether we read with Gratz and Gunkel,' ‘¢ Bless, O Jahweh, the tents of Shem,”’ or with Budde and Holzinger, ‘** Blessed of Jahweh is Shem,”’ there can be no doubt of the preference shown for Shem by the J group of writers to whom this blessing belongs. The source and original force associated with Shem? is as obscure as that of Ham. Back of both names no doubt lies a mass of traditions and possibly also myths which have been lost, but when once ‘in some way the favorable conception in regard to Shem had become current it would be natural for J to make the endeavor to trace the origin of his own people to this favorite son. Such is the purpose of the rather 1 See above, p. 181 seg. *Shem signifying “fame,” “ distinction,” has been compared with Aryan “ ruler,” “ noble,” as a designation of the favorite group (see Holzinger, Genests, p. 92), but all such explanations are open to the objection that they assume the introduction of the name to be due to the Hebrew writers, whereas it is evident that both Shem and Ham (which on the same supposition has been explained as * hot,” z.¢.,¢ the southerners’ of Holzinger, /, c.) are terms adopted by Hebrew writers and belong presumably to a much earlier age than their use in the Vé7- kertafel, There is much to be said ip favor of the view which regards both names as designations of old deities, though this view, likewise, is open to objec- tions which cannot easily be set aside, See Hommel, A/éisraelitische Ueberlie- Serung, pp. 47 and 115, 1903.) JASTROW—THE HAMITES AND SEMITES. 201 awkwardly constructed 21st verse of the roth chapter. The curi- ous phrase defining Shem as ‘‘ the father of all the sons of Eber’’ reveals the existence of an earlier tradition, which traced the Hebrews back to Eber. In view of this, one is tempted to conjec- ture that in an older form of the blessing at the end of the gth chapter, Eber took the place now occupied by Shem, so that the original personages concerned in the blessing and curse were Eber, Canaan and Japheth, subsequently enlarged to Shem, Ham and Japheth. However this may be, it is interesting and of some impor- tance to observe that when Eber was first associated with Shem, the former was made the son of the latter, whereas in the more scholastic ethnological scheme devised by P, the relationship of Eber to Shem was altered into that of greatgrandfather and greatgrandson.* How far this view already prevailed in pre-exilic days among some groups of writers it is, of course, impossible to say. Of the four® sons of Shem in P, Elam, Asshur, Arpachshad and Aram, it would appear that Elam is used as an inclusive term to embrace Babylo- nia. If this be correct we might have in this use of Elam an indi- cation of the date of P’s Vé/kertafe/, inasmuch as such a usage would point to the absorption of Babylonia by a power advancing from Elam. This power would, of course, be none other than Persia, and the use of Elam here as including Babylonia would thus force us to the conclusion that P’s list belongs to the close of the exilic period, subsequent to Cyrus’ conquest of Babylonia in 539 B.C. The theory, it must be admitted, encounters an obstacle in Arpachshad, if, as seems plausible, the latter embodies a reference to Chaldza, since it would involve the further supposition of a differentiation on the part of Hebrew writers between Chaldza and Babylonia. One can understand and indeed recognize the necessity of such a differentiation from the standpoint of one who, while placing Baby- 1Tt reads literally “and to Shem, there was born even he the father of all the sons’of Eber, the brother of Japheth the elder.” Then follows “the sons of Shem are Elam, Asshur, etc.” Comparing the beginning of verse 21 with the beginning of verse 25, “‘and to Eber were born two sons, etc.,” we should expect the enumeration of the sons of Shem immediately after the words “ and to Shem there was born.” *Shem, Arpachshad, Shelah, Eber (verse 24). ’Omitting Lud, which is a hopeless stumbling block (cf. Holzinger, Genesis, p. 105), and which as has above been suggested (p, 190) may have slipped in here through confusion with ya/ad in verse 24. 202 JASTROW—THE HAMITES AND SEMITES. [April 4, lonia and Assyria with the genealogy of Ham through Nimrod, the grandson of Ham, yet shares the common tradition which traces Eber the ancestor of the Hebrews (or Terah the descendant of Eber) to Ur-Kasdim, z.¢., to Chaldzea. Hence J, despite his hos- tility to Assyria and Babylonia, admits Arpachshad, which cer- tainly stands in some relationship to Ur-Kasdim, among the She- mites. Since P, however, places Asshur or Assyria with the sons of Shem, he does not share J’s view of Assyria or Babylonia, and there would be no reason why he should either omit Babylonia or specifically differentiate Chaldzea from Babylonia, unless it be, indeed, that he includes Arpachshad in obedience to the tradition which associated the latter with the home of his people. On the whole, this appears to be the more plausible view, for while P, as we have seen, manifests his purely scholastic interests to an astonishing degree, he yet is not entirely free from the natural spirit of national likes and dislikes, and at all events would be inclined to embody in his list current traditions regarding the origin of his people, even where such an embodiment might be superfluous or render his scheme somewhat ambiguous. Assuming then that Elam includes Babylonia, and that Arpachshad is Chaldzea, the Shemites, accord- ing to P, would represent the groups living in the district to which the Hebrews traced their origin, Elam, Babylonia, Assyria and Chal- dea, and the groups immediately to the west and northwest, classed by P under the general designation of Aram. We have no means of determining whether J’s list also included Aram among the sons of Shem, but there is also no positive evidence against it. . If it did, the genealogy of Aram was probably identical with the one pre- served in P, orat all events did not contain sufficiently important derivations to warrant the compiler who combined J with P in extracting anything from J’s list. So much seems certain, that J’s chief interest lay in Arpachshad, because of the supposed connec- tion between this district and Ur-Kasdim or Chaldza as the home of the Hebrews, and J’s interest here was sufficiently pronounced to induce him to carry down the line of Arpachshad in its two branches, Eber-Peleg and Eber-Joktan, the former representing a northern group, the latter a southern, much as the Arabs carry the genealogies of their clans to a northern and southern ancestor.’ ‘See Wistenfeld’s Genealogische Tabellen der Arabischen Stimme und Familien (Gottingen, 1852). 1903.] JASTROW—THE HAMITES AND SEMITES. 203 Of J’s double list, only the Eber-Joktan, the southern branch, has been preserved in the roth chapter,’ the northern branch being omitted by the compiler of J and P, because of its preservation in the P document in the 11th chapter, verses 16-26. The groups thus included in J’s genealogy of Shemites would be limited to those descended from Arpachshad and Aram, or since Arpachshad rep- resents on the one hand the Hebrews as the descendants of the north- ern branch of Eber-Peleg, and the Arabs on the other hand as the descendants of the southern branch of Eber-Joktan, the Shemites would be limited to groups in the immediate environment of the Hebrews. The point of view is apparently that of the Hebrew settlements to the east of the Jordan, Eber-Joktan representing the southern groups and Aram the northern and northwestern, with Eber-Peleg occupying the central position. Here too, therefore, we find J presenting a contrast to P, who, standing for an enlarged geographical and ethnological view, begins his enumeration with Elam to the East and passes on in a westerly and then northwesterly direction, which leads him to include Chaldza, Babylonia and Assyria and to end with Aram. The point of view here suggests Babylonia or Chaldzea as the home of P, or at all events as the central seat of the Shemites, with Elam constituting the eastern and Aram the western limit and environment. Consistent, moreover, with his view of the Hamites as the designation of groups settled in the remote south, he excludes those peoples included by J in the Eber- Joktan branch of Arpachshad from the Shemites, and as the Eber- Peleg list of P in the r1th chapter shows, P limits the Shemite branch of Arpachshad to the Eber-Peleg or northern division. The general scheme in P’s Vé/kertafe/ is thus quite clearly based on a geographical distribution into three zones, Japheth represent- ing the remote groups to the west, north and northeast, the Hamites representing the remote nations in the south, while the Shemites represent those in the immediate environment of the Hebrews from the point of view of a writer who, living in Babylonia, is influenced both by conditions prevailing in his days, by the tradition which traced the Hebrews to Chaldzea, as well as by the fact of the later settlements of the Hebrews to the east of the Jordan and in Pales- tine proper. Taking all these factors together, to which we ought perhaps to add the inclusion of Babylonia under Elam as due to 1 Verses 26-29. 204 JASTROW—THE HAMITES AND SEMITES. (April 4, special circumstances, the Shemites are for P the groups that live in districts in which, at one time or the other, the Hebrews had settled, or which represent districts adjacent to those settlements. The Shemites are, therefore, the groups that are ‘‘near’’ to the Hebrews as against the Japhethites and Hamites who are “ remote.’’ Again, while as we have seen P is actuated by large geographical views and displays considerable ethnological knowledge set forth in a scholastic spirit, it is natural that when he comes to the group to which his own people belongs he should show some traces also of a nationalistic spirit. His general point of view in regard to the She- mites as representing those nations which are adjacent to the Hebrews, or ‘‘near’’ them, may be put down to the credit of his nationalistic - spirit, while the departure from his scheme in including Canaan among the Hamites instead of placing them with the ‘ near’”’ nations or Shemites may represent a trace of the influence of the spirit of hostility towards Canaanites which controls J, though it is also possible that the addition of Canaan in verse 6 is due to the compiler who combined J with P, and who is actuated by the same spirit as is J. X. In sharp contrast, both as to geographical views and ethnographi- cal knowledge and general spirit, stands J and the group of writers to which he belongs or who follow in his path. Showing distinct traces of the older view which limits the geographical horizon to three groups, Shem (or perhaps Eber’), Canaan and Japheth, all dwelling within the confines of Hebrew settlements in Palestine, J, though representing an enlarged view in substituting Ham for Canaan (and Shem for Eber), and in extending Japheth to include remote nations with which Hebrew history has nothing to do, arranges his Vé/kertafel entirely from the Hebrew point of view. Though apparently agreeing with P in his definition of Japhethites it is doubtful whether J’s list of sons and descendants of Japheth was as extensive as P’s, and at all events the Japhethites did not represent the geographically remote nations but rather those that were historically remote, toward which a writer interested primarily ‘In view of the importance which Eber plays as the ancestor of the two groups, Eber-Peleg and Eber-Joktan, it would indeed appear as already suggested (p. 201) that a tradition was current which made him rather than Shem the an- cestor of the groups allied with the Hebrews. 1903.] JASTROW—THE HAMITES AND SEMITES. 205 and, indeed, exclusively in Hebrew history would be wholly indifferent. ‘The Japhethites are, accordingly, no longer ‘‘ a group dwelling in the tents of Shem,’’ but quite outside of these tents. More marked is the contrast between J and P in regard to the Hamites. While here, too, it happens that from P’s point of view Egypt falls into the category of the Hamites, still the whole char- acter of Ham’s genealogy in P shows that this is done because of the agreement with P’s definition of Hamites as embracing the “‘remote’’ nations of the south. In the mind of J, however, the Hamites take the place of Canaan, the ‘‘ accursed ’’ son of Noah, and the enlargement of Canaan to Ham furnishes him with the opportunity of adding to Canaan a whole series ef nations who, because of the mischief they wrought at one time or the other in Hebrew history, merit the fate of being cast into the purgatory of the ‘‘accursed’’ nations. From this point of view, J includes Egypt among the Hamites and adds to Canaan and Egypt, the Baby- lonians and Assyrians, as well as the Philistines, while subsequent writers, actuated by the same spirit as J, are at pains to specify the subdivisions of the Canaanites, and with a view of leaving no pos- sible loophole for such types of ‘* wickedness ’’ as Sodom and Gomorrah, even indicate the exact boundaries of the Canaanitish settlements. The mention of Sodom and Gomorrah, even if the view above set forth that the names are intended to include Moab and Ammon be not accepted, shows too clearly to admit of any doubt the picture in J’s mind of the character and nature of the Hamites. Coming to Shem, there is a closer approach to be observed be- tween the views of J and P and yet even here there are some strik- ing contrasts. Not only is P’s list of Shemites on the whole more inclusive, since he makes them extend from Elam in the East to Aram and Palestine in the West, though on the other hand he ex- cludes the southern Arabs who in J represent the southern branch of Eber-Joktan, but his historical standpoint is also larger than that of J, since he embodies in his list not only the tradition of the original home of the Hebrews but draws the proper conclusion from this tradition that the inclusion of Arpachshad or Chaldza among the sons of Shem carries with it Babylonia (including Elam) and Assyria. J in all probabilities included Aram by the side of Arpachshad among the sons of Shem, but his point of view is that of one who is exclusively interested in Hebrew history. The im- 206 JASTROW—THE HAMITES AND SEMITES. [April 4, portance of Shem lies for him in the fact that Shem is ‘‘ the father of all the children of Eber.’’ For him the ‘‘ remote ’’ nations of the south, with whom Hebrew history is as little concerhed—barring ‘the relationship between Judzea and southern Arabia reflected in the legends of King Solomon’s dealings with the Queen of Sheba’— as the ‘‘remote’’ nations of the north and east represented by the Japhethites, are not as in P the Hamites, but the groups which represent the subdivisions of southern Arabia. J, therefore, like P has two ‘‘remote”’ groups, but the entire character of the former’s Vilkertafel is changed by his definition of the Hamites as those representing the enemies of Israel. To sum up, therefore, J’s list includes three general groups: (@) peoples towards whom J was indifferent because of little or of no moment to Hebrew history, (4) peoples towards whom he harbored bitter feelings of hostility, and (c) his own people towards whom he was partial and whose descent he traced, from the favorite son of Noah. The first group includes again (1) the Japhethites in the west, north and northeast and (2) the Eber-Joktan branch or southern Arabs. His nationalistic spirit manifests itself in those whom he places in the second group, while on the other hand the limits to this spirit are represented by his willingness to place the: southern Arabs among the favored Shemites, being prompted to this display of generosity by the absence of any motive for excluding them and the self-evident consideration that the Shemites must include other subdivisions besides his favorites—the Hebrews. The scholastic spirit which J also possesses when it does not interfere with his natural dislikes, though not to the same degree as P, leads him likewise to recognize the close relationship between Hebrews and Arameans, so that his Shemites as seems likely also included Aram. P, on the other hand, free from the nationalistic spirit, except pos- sibly when the Canaanites are involved, sets up two very well-defined groups; (a) the remote nations of the west, north and northeast— the Japhethites, and (4) the remote nations of the south—Nubia, Egypt and southern Arabia—the Hamites, to which he adds (¢) as a third group the Hebrews and those adjacent or ‘‘near’’ to them, though his definition of ‘ near’’ again displays a larger historical ' I Kings, chap, x—amplified by further details in the “ Midrashic” litera- ture, See ¢. x. Weil's Biblische Legenden der Muselmanner, pp. 247-271. 1904. | OLIVER—COLOR-SIGNALS IN MARINE SERVICE. 207 and geographical view than does J and the school of writers that follow him. Lastly, it is to be noted that the writers responsible for the numerous additions and glosses to J as well as the compiler who combined J with P stand under the influence of the narrower view manifested by J, so that in its present form the Vé/kertafe/ in the*tenth chapter of Genesis regarded as a ‘‘ combined ’’ document impresses one as bearing out J’s conception of Hamites and Shemites, the former as the ‘‘accursed,” the latter as those ‘* blessed ’’ by Jahweh, rather than P’s definition. Nor is it sur- prising, in view of political events and religious developments in the post-exilic period, that the more rigidly ‘‘scholastic’’ division of nations should have been eclipsed by one that appealed more to the national interests and that must have been a source of hope and consolation in trying days—encouraging the Hebrews to look for- ward to a time when the ‘‘curse”’ and ‘‘ blessing’’ pronounced on Ham and Shem, or Canaan and Eber, respectively, would be ful- filled. University of Pennsylvania, June, 1904. REGULATION OF COLOR-SIGNALS IN MARINE AND NAVAL SERVICE. BY CHARLES A. OLIVER. (Read April 9, 1904.) When it is considered that the most dangerous periods of time for the safety of lives and preservation of property at sea are those during which the proper recognition of color-signals constitutes the main and, at times, the only guide for immediate action, the impor- tance of the regulation of the choice of the colors used, the character of the materials employed, the size of the objects submitted for inspection, and the degree and the character of the visual acuity necessary for the determination cf such colors, become evident. So long as the high seas are necessarily free, and harbors con- stantly changing in topography and ofttimes difficulty of access ; rivers and streams occupied in similar places by crafts of varied size and differing speeds ; permanently fixed objects, such as buoys and direction and danger indicators, must have color differentiation employed as their main expressive feature ; and color-signs must be used to signify the position of large floating masses, such as ships at 208 OLIVER-—COLOR-SIGNALS IN MARINE SERVICE. [April 9, anchor,—just so long will it remain necessary continually to improve the color material employed during actual service, and to _ render the apparatus which is to be used the most simple in con- struction that can be employed. The well-filled harbor, with its changeable and constantly cross- ing paths containing traffic of every conceivable kind, the insta- bility of the water mass itself, and the uncertain factors, such as fogs, mists and snow, all show to what great degree of danger every moving object placed within such a situation is exposed. These conditions are far different in degree of uncertainty from those that are seen in railway travel, in which the directions of movement are comparatively fixed, every change of direction well protected, and all of the trains carefully guarded by block systems. The first question which arises is, Can the system of signalling now in vogue in marine and naval service be so changed as to give better results with less liability to error?! Experiment and trial have shown that the visual apparatus which projects man’s ordinary sensory powers possibly to the greatest distance into space must be the sensory organ which is preferably to be employed during the common routine of duty. Fixed or intentionally changed color differentiations being less unstable, and hence more certain for visual perception than mere recognition of form and objective motion, must be that which should be practically employed. As the result of experience, the coarse colors, red, green, yellow, white and blue, are the ones which have been found to be the best for use during maritime signalling. These colors which are either placed in related situations upon movable bodies (both while in motion and while at rest upon bodies of water), or which are situated in fixed positions, are made interchangeable and time-regulated. These colors, possessing definite color-arrangement and color-sequence, are intended either to express direction, signify protection or designate code-signalling ; varieties of work—the correct and, at times, vital answers to which are dependent solely upon color recognition at distances which are ' Better, less complicated, and hence cheaper and more easily applied adapt- ations of the Hertzian Ray apparatus might accomplish the purpose in one way; but unfortunately, unless such instrumentation is automatic in action, and unless its management and use can be kept constantly correct, this method must be considered in the light of the future. 1904.] OLIVER—COLOR-SIGNALS IN MARINE SERVICE. 209 comparable with safety to large moving masses that often can be alone stopped slowly and gradually—colors and relative positions which must be carefully chosen in regard to distances, situ- ations, etc. In the following paragraphs it has been endeavored to express clearly and briefly the specific reasons for the improvements and changes suggested. I. All of the color tints to be used both by reflected light- stimuli and transmitted light-stimuli (day and night) during actual duty, should be officially proven copies of standards which have been carefully chosen in such a way that the signals may be uniform in tint in spite of variations in the character of the illuminants themselves. These selections should be made by an international commission of normal-eyed color experts. The color-signals will then be universally alike, thus minimizing danger from confusion due to false color exposure. These results can probably best be obtained by mathematically and analytically obtaining sample pigment hues, both for diffuse reflected solar light and diffuse refracted artificial light, of specified kinds, character, degrees and tints, which are equivalent to the midway bands for the colors used in the corresponding portions of the color spectra obtained during exposure to the illuminant to be employed during actual service. II. Each vessel of any importance should be provided with pro- portionately-sized miniature samples of color-boxes, color-lamps, signal-colors, etc.,—or better, fitted with full-sized examples of the same,—all carefully protected and boxed. These should be used as guides for the tinting of all material which employs color as its basis for signalling of any kind. These materials should be certified by proper authority, and should be obtainable at cost at licensed shops in every port of any consequence. III. It should be a part of the official duty of every national, state and municipal government to see that the materials which are used for color-signalling in any form, as well as the samples, are periodically examined as to cleanliness and stability of tint. Dated certificates, brief and to the point, with plain instructions for the easiest manufacture and the best plans for the preservation of the color materials, together with clearly expressed rules for dis- tances used, situations employed, and notes on any color peculiarity PROC. AMER. PHILOS. SOC. XLII. 176. N. PRINTED JUNE 27, 1904, 210 OLIVER—COLOR-SIGNALS IN MARINE SERVICE. [April 9, of certain places, should be given; these to be submitted for inspection on demand. IV. Every series of related colors used should be regulated, both as to their comparative sizes of exposure and the relative degrees of color saturation ; these should be duly proportionate in reference to equalities, distinctness, relationships and association of safe dis- tances, and with regard to differences in degrees of penetrability. This can be accomplished either by having the color values graded proportionately (a bad plan, since it tends to weaken the value of the stronger colors), or by making the color areas relative in size: for example, to give a green signal light a similar degree of brightness, and hence the same relative distinctness (which governs all apparent distances, and in consequence the relationships of the two colors), as red, it must either be five times more power- fully illuminated than the red or given five times more exposed superficial area: so too with all other color changes; there is an idiocratic relationship. Clinical experiment has shown this, and laboratory research has confirmed the practical findings. The importance of this factor can hardly be overestimated when the series of individual signal colors are numerous in well-filled and busy harbors. These plans once agreed upon by such an international com- mission, all necessary data will soon become common property, and in consequence the system will be universally understood. Philadelphia, April 7, 1904+ 1904.] BREZINA—COLLECTIONS OF METEORITES. 211 THE ARRANGEMENT OF COLLECTIONS OF METEORITES. BY DR. ARISTIDES BREZINA, VIENNA. Late Director of the Mineralogical Department of the Natural History Museum at Vienna. (Read April 8, 1904.) In making a collection of any kind of matter two ends must be kept in view; firstly, to secure in due time and to preserve as great and complete a variety of the material as possible, and secondly, to illustrate as fully as possible all ways in which the matter may be considered. According as a collection provides for the first or the second purpose it is called a systematic or a synoptical collection. Until 1889 there existed meteorite collections of the first kind only; in this year the new Natural History Museum of the Court at Vienna was to be opened; and as for a hundred years this collection had been worked upon by Chladnt, Schretbers, Widmanstitten, Partsch, Haidinger, Hérnes, Wohler, Tschermak and myself, most of its specimens from different localities had been investigated structurally and chemically so thoroughly, that I could for the first time divide the material into ten great series. They were disposed as follows: I. Ancient coins on which sacred meteorites were repre- sented. II. Historical meteorites which were worshiped by primi- tive nations or which formed standards in the development of meteoric science; related bodies, as fallen dust, bloody rain, meteor-paper, nuclei of hail and pseudo-meteorites. III. Specimens of meteorites which show processes of melt- ing, incrustation, cleaving and faulting, black and metallic veins, etc., (celestial alterations); and the results of experi- ments on meteorites for producing similar alterations. IV. Specimens showing terrestrial alterations, viz., defor- mation by falling on the earth, erosion of the surface by terrestrial agents, chemical alteration after the fall, forma- tion of new constituents by humidity, etc. V. The constituents of the meteorites,,from simple minerals * 212 BREZINA—COLLECTIONS OF METEORITES. [April & to complex bodies, free and in combination; artificial pro- ducts, meteorites and their minerals formed synthetically. VI. Slices through whole meteorites from all petrographic groups, showing the general structure on large surfaces. VII. A series of specimens of equal size and normal con- stitution of all groups of meteorites, free from extraordinary inclusions, but showing the differences between pieces of the same fall; this series was intended to allow of a quick deter- mination of the group to which a new meteorite belongs. VIII. A collection of microscopic slides of all meteorites fit for microscopic study. IX. The systematic collection, containing the main mass of the collection, arranged from the petrographical and minera- logical points of view. X. Casts of all meteorites of characteristic outer form. _A collector who has considerable means at his disposal should begin by forming a systematic collection. An excel- lent example of such a collection built up within a moderate number of years is the Ward-Coonley collection,-whose cata- logue has lately been published. It represents a greater number of localities than any other public or private collec- tion in the world, nearly 90 per cent. of all meteorites known; and it averages so high in weights, that later it may from its surplus furnish the material for all kinds of researches and for exchanges on the largest scale, so that it will be indepen- dent from acquisitions by purchases for a long time and will permit of the formation by and by of a synoptical collection. In the following pages I give the description of a collection of this second kind, which I have formed since 1896, derived from a small number of monopolized falls by numerous ex- changes. This description may serve as a guide for collectors who intend to develop a synoptical collection out of the material of a systematic one. I. Betyt Corns. The ancients supposed the stars to be the domiciles of the gods; falling stars and falling meteorites signified the descending of a god or the sending of its image to earth. 1904.) | BREZINA—COLLECTIONS OF METEORITES. 213 These envoys were received with divine honors, embalmed and draped and worshiped in temples built for them. From about 400 (or 500) B.C. to 300 A.D. coins were struck in honor of these divinities by emperors and autonomous cities. In general the images were rather naturalistic in older times and became human-like (iconic) afterward. Many of these betyl coins represent stones reported ex- pressly to have fallen from heaven; some of them present as a common feature the likeness to conic stones, or obelisks or to archaic, half-iconic simulacra: so it comes that similar representations of unknown origin were likewise supposed to represent sacred meteorites. A. Betyls Reported to Have Fallen from Heaven. 1. The Omphalos of Delphi.—A black stone which was given by Rhea to Uranos instead of the new-born Zeus, and ren- dered to Zeus after his victory over Kronos; Zeus or Saturn threw it on the Earth, on the point which was considered as the centre of the Earth. Eleuthernai, Kreta; Autonomous 2 AE.! Makedonia; Philippus II AR. PI. I, Fig. 1. Myrina, Aiolis; Auton. AR. Fig. 2. Nakrasa, Lydia; Auton. AE. Fig. 3. Neapolis, Campania; Auton. 3 AE. Fig. 4. Parthia; Tiridates 3 AR, Fig. 5, Arsaces II AR, Phriapa- tius AR, Phraates I 4 AR, MithradatesI 5 AR. Fig. 6. Roma, Italia; Sabina AE. Syria; Antiochus I Soter ro AR, Fig. 7, Antiochus II Theos AR, Antiochus Hierax 2 AR, Seleucus III Ceraunus 3 AR, Antiochus filius AR, Antiochus III Magnus 17 AR, Seleucus IV Philopator 2 AR, Antiochus IV Epiphanes AR, Demetrius I Soter 3 AR, Fig. 8, Alexander I Bala 4 AR. 2. The Stone of Emisa, El Gabal.—A black, conical stone, which Herodian reports to have fallen from heaven; Elagabal transferred it to Rome, where it remained until 222 A.D. 1 AE bronzes, AR argent, AV aurum (gold); the number before gives the number of different kinds represented: 2 AE, two bronzes. 214 BREZINA—COLLECTIONS OF METEORITES. [April 8, - Emisa, Seleukis and Pieria; Antoninus Pius 2 AE. Fig. 9. Roma, Italia; Elagabalus 3 AR. Fig. ro. 3. Zeus Katatbates—The descended god, who was repre- sented sitting on a throne. Kyrrhos, Kyrrhestika; Trajan AE, Fig. 11, Antoninus Pius 2 AE, Marcus Aurelius 2 AE, Lucius Verus AE, Commo- dus AE, Elagabal AE, Philippus pater AE, Fig. 12, Philippus filius 2 AE. 4. Aphrodite Paphia.—A stone said to have fallen from heaven as an image of the Paphian Aphrodite; an elongated cone in a temple of two columns. Kypros; Galba AE, Vespasian 4 AR, Fig. 13, AE, Trajan AE, Septimius Severus AE, Julia Domna AE, Caracalla AE. Fig. 14. 5. Artemis Ephesia.—Image of Artemis reported to have fallen from heaven and been preserved in the temple of Ephesos. Form half-iconic. Aizanis, Phrygia; Auton. AE,Commodus AE. PI. II, Fig. 15. Ankyra, Phrygia; Sabina AE, Faustina jun. 2 AE, Julia Domna AE. Ephesos, Ionia; Antoninus Pius AE. Ephesos and Pergamon; Commodus AE, Gallienus 2 AE. Figs. 16, 17. Eumeneia, Phrygia; Auton. AE. Nakrasa, Lydia; Auton. 2 AE. Philadelphia, Lydia; Auton. AE. Provincia Asia; Claudius AR, Fig. 18, Hadrian 2 AR. - Roma, Italia; Hostilia AR. Fig. 19. Tabai, Karia; Auton. AR. Tiberiopolis, Phrygia; Trajan AE, 6. Stone of Astarte, which fell as a star from heaven and was raised by Astarte, who consecrated it to the town of Tyros; a second stone of Astarte was worshiped at Sidon | (represented laying on a car), and some coins of Tyros exhibit both stones (the Ambrosian petra). Sidon, Phoinikia; Auton. 3 AE, Hadrian AE, Caracalla AE, Elagabal 9 AE, Fig. 20, Julia Soemias AE, Julia Mesa AE, Annia Faustina AE, Alexander Sever AE. 1904,] BREZINA—COLLECTIONS OF METEORITES. 215 Tyros, Phoinikia; Elagabal AE, Fig. 21, Valerianus pater AE, Salonina AE. B. Betyls Accepted by Analogy to Represent Meteorites. 7. The Pyramids or Obelisks of Apollon.— Ambrakia, Epeiros; Auton. 3 AE. Apollonia, Illyria; Auton. 3 AE. Fig. 22. Megara, Megaris; Auton. AE. Myrina, Aiolis; Auton. AE. 8. The Herms of Hermes, Ithyphallos and Priapos.— Makedonia; Alexander III Magnus AR. Fig. 23. 9g. Telesphoros.— Akrasa, Lydia; Auton. AE. Eukarpeia, Phrygia; Auton. AE. Fig. 24. Hadrianeia, Mysia; Antoninus Pius AE. Roma, Italia; Caracalla AE. 10. The Two Stones of Zeus Dolichenos or Herakles Sandan.— Syria (Tarsos); Antiochus VIII Grypus 3 AR, Fig. 25, Antiochus IX Cyzicenus AR. Tarsos, Kilikia; Auton. 6 AE. 11. Zeus Kasios, represented as a conical stone suspended by a chain in a strap. Seleukeia, Seleukis and Pieria; Trajan 8 AE, Fig. 26, Antoninus Pius 2 AE, Marcus Aurelius AE, Commodus AE, Septimius Severus AE, Caracalla AE, Alexander Severus AE. 12. Conical or Quadratic Stones without determination.— Mallos, Kilikia or Rhosos, Seleukis and Pieria; Auton. 2 AR. Fig. 27. Perga, Pamphylia; Gallienus AE. Synnada, Phrygia; Gallienus AE. Fig. 28. 13. The Conical Stone of Aphrodite Urania.— Makedonia (Uranopolis); Alexander III Magnus 12 AR. Pl. III, Fig. 29. Uranopolis, Chalkidike; Auton. AE. 14. The Simulacrum of Artemis Anaitis, half-iconic.— Apameia, Phrygia; Auton. AE. Hypaipa, Lydia; Trajan AE. Fig. 30. 15. The Simulacrum of Artemis Leukophrys, half-iconic.— Magnesia ad Maandrum, Ionia; Auton. AE. Fig. 31. 216 BREZINA—COLLECTIONS OF METEORITES. [April 8, 16. The Simulacrum of Artemis Pergaia; a cone with human head.— Kaisareia, Kappadokia (Provincia Asia); Trajan AR. Fig. 32. Perga, Pamphylia; Auton. 2 AE, Trajan AE, Caracalla AE, Diadumenian AE, Tranquillina 2 AE, Fig. 33, Philippus pater 3 AE, Valerianus pater AE, Gallienus 3 AE, Aurelian AE. Pogla, Pisidia; Antoninus Pius AE. Provincia Asia; Nerva 2 AR, Trajan3 AR. Fig. 34. 17. The Simulacrum of Astarte, half-iconic.— Gabala, Seleukis and Pieria; Macrinus AE. 18. The Conical Stone of Hera.— Chalkis, Euboia; Auton. AE. Fig. 35. 19. The Simulacrum of the Samian Hera, half-iconic.— Panionion, Ionia; Marcus Aurelius AE. Fig. 36. Samos, Ionia; Caracalla AE, Fig. 37, Alexander Severus 3 AE, Philippus pater 2 AE, Trajanus Decius 2 AE, Etruscilla 2 AE, Gallienus AE, Fig. 38. 20. The Simulacrum.of Persephone, half-iconic.— Julia Gordos, Lydia; Marcus Aurelius AE. Sardes, Lydia; Auton. AE, Fig. 39, Salonina 2 AE. Sardes and Hierapolis; Philippus pater AE. 21. Archaic Simulacrum of Double Goddess.— Capua, Campania; Auton. AE. C. Related Celestial Bodies. 22. Comets.— Roma, Italia; Sanguinia (Julius Cesar) AR, Augustus 5 AR. Fig. 40. Silesia, Germany (modern); Auton. AV, Fig. 41, 2 AR, AE, II. HistoricAL METEORITES. Mordvinovka, Ekaterinoslav, Russia. Prehistoric. Cw.' I gram, 2 cm. Casas Grandes, Mexico. Prehistoric. Om. 102 gr. 26cm. These two meteorites were found in prehistoric tumuli, The symbols following the date of fall designate the petrographic group as defined later on (VII, System of Meteorites), Weight and profitable surface are added in grams and square centimeters. 1904.] BREZINA—COLLECTIONS OF METEORITES. 217 Wichita county, Brazos, Texas. Found 1836. Og. 168 gr. 44 cm. Worshiped by Indians as coming from ‘‘the Great Spirit.”’ Kesen, Iwate, Japan. Fell June 13, 1850. Ccb. 8 gr. 5 cm. Worshiped in a temple of Iwate as a bety]. Ensisheim, Alsace, Germany. Fell November 16, 1492. Ckb. 1ogr.5 cm. Oldest meteorite of known fall. Elbogen, Bohemia. Known since about 1400. Om. 12 gr. 4 cm. The so-called ‘‘verwunschene Burggraf.”’ Said to have occided in 1410 the burggrave Botho von Eulenburg, . defamed for his cruelty, the ancestor of the principles of Eulenburg. ! Krasnojarsk, Siberia. Found 1749. Pk. 25 gr. 8 cm. The meteorite on which Chladni demonstrated the cosmic nature of these bodies. Albareto, Modena, Italy. Fell July, 1766. Cc. 3 gr. 2cm. The fall was described carefully by the Jesuit Troili in the paper ‘‘Della caduta di un sasso dell’ aria.” Modena, 1766, 128 pages. Barbotan, France. Fell July 24,1790. Cga. 9 gr. 4 cm. These stones were thrown away by the Paris Academicians, who feared the ridiculousness, if believing in the reality of the fact of falling stars. Laigle, France. Fell April 26, 1804. Cib. 115 gr. 18 cm. The fall was examined carefully by the celebrated Biot and dissipated the doubts in France. Borodino, Russia. Fell September 5-6, 1812, during the battle of Borodino, in the chief-quarter of the Russians. Cgb. 5 gr. 2cm. Mazapil, Mexico. Fell November 27, 1885, with the Bielid star-shower which replaced the disappeared Biela comet. Om. 11griscm. | Bjurbdle, Finland. Fell March 12, 1899, in the bottom of Stensbéle Fjérde, and was raised by divers; shows adhering sea-ooze. Cca. 61 gr. 12 cm. Breslau, Silesia. Fell January 31, 1848. 4 gr.—Rescht, Persia. Fell September 10, 1864. 3 gr—Nacimiento del Rio Colorado, Argentina. Fell 1883. 1 gr.—Meteoric dust containing nickel. 218 BREZINA—COLLECTIONS OF METEORITES. [April 8, Neusohl, Hungary. Fell January, 1848. 0.3 gr.—Nieder- tenzel, Bohemia. Fell February 18, 1899. 11 gr.—Terres- trial dust containing no nickel. é Ivan, Oedenburg. Hungary. Limonite pebbles which fell August 10, 1841, in accumulations of hundreds of tons, being raised by a cyclone from the exsiccated grounds of Lake Neusiedel. III. _ScaTTERING OF METEORITES. Brenham, Kansas. Found 1885. Pk. Free of olivine. 250 gr. 38 cm.— Brenham, Kansas. Found 1885. Pk. Rich in olivine. 109 gr. 45 cm.—Glorietta, New Mexico. Found 1884. Pk. Free of olivine. 119 gr. 36 cm.—Members of the chain-fall Brenham (37° 38’ N., 99° 13’ W.), Glorietta (35° 52’ N., 105° 30’ W.), Port Orford (42° 46’ N., 123° 10’ W.). Lerici, Italy. Fell January 30,1868, 7 p.m. Cga. 6 gr. 3 cm.—Pultusk, Russia. Fell January 30, 1868, 7 p.m. Cgb. 57 gr.—Alike in structure, fell at the same time on a line coinciding with the flying-line of Pultusk; Lerici 44° 4’ N., 9° 55’ O., Pultusk 52° 42 N., 21° 23’ O. Vaca Muerta, Chile. Known 1861. Mg. 30 gr. 11 cm.— Cerro la Bomba, Taltal, Chile. Found 1888. Mg. 151 gr. 18 cm.—Quebrada Huanilla, Cachinal, Chile. Found 1887. Mg. 2 gr. 2 cm.—Mejillones, Chile. Found 1867. Mg. 1 gr. 1 cm.—Pieces misplaced by rancheros. Lion River, Great Namaland. Found before 1853. Of. 27 gr. 7 cm.—Bethany-Berseba, Namaland. Known in 1874. Of. 4 gr. 3 cm. Mukerop, Gibeon, first block. Found in - 1899. Of. 500 gr. 42 cm.—Mukerop, Gibeon, second block. Known in 1902. Of. 673 gr. 36 cm. Lion River circa 23° 40’ S., 17° 40’ W.; Bethany 26° 30’ S., 16° 49’ W.; Berseba 26° 0’ S., 17° 42’ W.; Gibeon 25° 6’ N., 17° 48’ W. IV. MELTING AND Fusion, ScoRIFICATION, FAULTING, SEPA- RATING, Stannern, Moravia, Fell May 22, 1808, Eu. 162 gr. 28cm, Crust melted easily; fritted earth on front side. 1904,] BREZINA—COLLECTIONS OF METEORITES. 219 Juvinas, France. Fell June 15, 1821. Eu. 21 gr. 9 cm. Crust melted easily, with small rolls. Mécs, Hungary. Fell February 3, 1882. Cwa. 135 gr. Five individuals; crust thick, scorified. Orvinio, Italy. Fell August 31,1872. Co. 30gr. Whole individual orientated, scorified crust, interrupted. Kernouvé, France. Fell May 22, 1869. Cka. 173 gr. 42cm. Crust loose, scabby, partly fallen off. Antifona, Italy. Fell February 3, 1890. Cc. 241 gr. 42 em. Apex of the stone with radial drift. Aleppo, Turkey. Found 1873. Cwhb. 67 gr. 15 cm. Crust of back, scoriated and bubbly. Pultusk, Russia. Fell January 30, 1868. Cgb. 233 gr. 30 cm. Whole individual, highly orientated; front crust primary, back crust secondary, thin. Knyahinya, Hungary. Fell June 9, 1866. Cg. 292 gr. so cm. Highly orientated whole individual, front drift, roll border; back crust thin, brown, crust-sprinkled. Knyahinya. 127 gr.18 cm. Individual of loaf form with uncovered striking spots. Estherville, lowa. Fell May 10,1879. M. 4ogr.6and 3 em. Two whole individuals, the one with metallic, the other with stony crust. Marjalathi, Finland. Fell June 1, 1r902. Pi. 147 gr. 22 em. Black crust, fine-drusy and even over the iron, even and bright over the olivine. Glorietta, New Mexico. Found1884. Pk. 492 gr. 40cm. Highly orientated individual, free of olivine; front drift, roll border, back crust loose, partly fallen off. Pultusk, Russia. Fell January 30, 1868. Cgb. 252 gr. 30 cm. Whole individual, polyhedral flake with primary crust. Pultusk, 154 gr. Thirty-eight whole individuals of equal weight (4 grams each) with primary and secondary crust. Hessle, Sweden. Fell January 1, 1869. Cc. 10 gr. 4 cm. Whole individual with four primary faces; apex-concavity after chrondule has fallen out. Hessle. 13 gr.5 cm. . Individual of flake form. Kansada, Kansas. Found 1897. Cib. 135 gr. 22 cm. 220 BREZINA—COLLECTIONS OF METEORITES. [April &,. Whole individual, primary faces, partly even, partly piezo- glyptic. Forest, Iowa. Fell May 2, 1890. Ccb. 83 gr. Five indi- viduals with faces partly polyhedral, partly rounded. Cafion Diablo, New Mexico. Found 1891. Og. 122 gr. 28 cm. Individual in form of acute-angular, piezoglyptic flake. Cafion Diablo. 51 gr. Twelve individuals of similar sharp angular flake form. Aumiéres, France. Fell June 3, 1842. Cwa. 18 gr. 5 cm. Flawy crust (craquelé) on concave face. Dhurmsala, India. Fell July 14,1860. Ci. 257 gr. 40cm. Stone which was extremely cold when it reached the earth. ~ Ochansk, Russia. Fell August 30, 1887. Ccb. 12 gr. 5 em. Infiltration of crust and crust-drops on uncovered fis- sure, issuing from bubbly scoriaceous crust. Girgenti, Italy. Fell February 10, 1853. Cwa. 132 gr. 22 cm. Parallel infiltration-veins with lateral apophyses. Plate IV, Fig. 42. Fisher, Minnesota. Fell April 9, 1894. Cia. 208 gr. 27 cm. Ramified crust-infiltration. Maémé, Japan. Fell November 10, 1886. Cia. 97 gr. 24cm. Chondrule transpierced by infiltration-vein. Aumiéres, France. Fell June 3, 1842. Cwa. 18 gr. 5 cm. Chondrule faulted by a black vein. Baratta, New South Wales. Fell May, 1845. Cgb. 45 gr. 18 cm. Broken chondrule, the parts dislocated. Chéteau-Renard, France. Fell June 12,1841. Cia. 80 gr. 12cm. Thick harness uncovered. Stilldalen, Sweden. Fell June 28, 1870. Cgb. 34 gr. g cm. Harnesses in different directions, partly uncovered. Alessandria, Italy. Fell February 2, 1860. Cga. 15 gr. scm. Harnesses and black crust-veins. Zavid, Bosnia. Fell August 1, 1897. Cia. 310 gr. 31 cm, Thin harness on even rupture-face, Lasdany, Russia. Fell June12,1820, Cga. 77 gr. 12cm. Harnesses on uneven rupture-faces. Badger, New Mexico. Known1897. Om. 427 gr. 45 cm. 1904.] BREZINA—COLLECTIONS OF METEORITES. 221 Empty octahedral fissures, crust in an excavation, stowing of lamelle. Badger. 461 gr. 30 cm. Octahedral fissures filled with magnetite; faulting and bending of lamelle. Chantonay, France. Fell August 5, 1812. Cgb. 89 gr. 25 cm. Fluidal structure of black parts. Mocs, Hungary. Fell February 3, 1882. Cwa. 6 gr. Unchanged original mass and pieces heated in copper enclosure (blackened) Knyahinya, Hungary. Fell June 9, 1866. -Cg. 4 gr. Heated in copper enclosure and blackened. Albareto, Italy. Fell July, 1766. Cc. Ignition prepara- tion (from old Italian experiments). Arlington, Minnesota. Found 1894. Om. 28 gr. 6 cm. Tron-enamel, alteration-zone along natural surface. Ngoureyma, Algiers. Fell June 15, 1900. Obzg. 29 gr. “cm. Iron-slag in fissure; molten and tracted mass. Carlton, Texas. Found 1887. Off. 147 gr. 30cm. Iron- slag in concavity; bending and faulting of lamelle. Jamestown, Dakota. Found 1885. Of. 69 gr. 20 cm. Melting-slag and alteration-zone. Mukerop (Bethany), Namaland. Found 1899 (1853). Of. 350 gr. 30cm. Melting-slag and alteration-zone. Reed City, Michigan. Found 1897. Oge. 122 gr. 23 cm. Mollified; alteration-zone. Hammersley, Australia. Found 1894. Om. 119 gr. 22 cm. Alteration-zone of 60 mm. thickness. Silver Crown, Wyoming. Found 1887. Og. 134 gr. 35 cm. Alteration-zone of 3-4 mm, thickness. Plate IV, Fig. 43. Sarepta, Russia. Found 1854. Og. 1g9gr.gcem. Altera- tion-zone, inner curve equalized, 1-3 mm. thickness. Barranca Blanca, Chile. Found 1885. Obz. 17 gr. 4 cm. Alteration-zone of 1-3 mm. thickness. Oscuro Mountain, New Mexico. Known 1898. Og. 67 gr. 17cm. Alteration-zone of o.4-2 mm. thickness. Ballinoo, Australia. Found 1893. Off. 395 gr. 40 cm. Double alteration-zone, outer zone sparkling, inner dull. Plate V, Fig. 44. Azucar, Chile. Found 1887. Og. 160 gr.35 cm. Alter- 222 BREZINA—COLLECTIONS OF METEORITES. [April ation-zone of 1-8 mm. thickness, terminating at a concavity produced by molten and removed Troilite. Puquios, Chile. Found 1894. 70 gr. 12 cm. Faulting of octahedral lamelle. Bridgewater, North Carolina. Described 1890. Om. 128 gr.30cm. Faulting of octahedral lamelle. Mukerop (Bethany), Namaland. Found 1899 (1853). 386 gr.35 cm. Faulting of lamelle on border-fissures. Mukerop. 64 gr.g cm. Wall-border bent on two sides. Mukerop. 400 gr. 42 cm. Wall-border bent on one side. Bella Roca, Mexico. Described 1888. Of. 104 gr. 15 cm. Wall-border bent on one side. Mukerop. 727 gr. 38 cm. Hexahedral chamfers. Lime Creek, Alabama. Found 1834. H. 8 gr. 2 cm. Neumann-lines bent. DeSotoville, Alabama. Found 1878. H. 158 gr. 25 cm. Canted Giant-Rhabdites on curveted faulting vein. Badger, New Mexico. Known1897. Om. 162 gr. 52 cm. Strong bending of inner octahedral lamelle. Smith Mountain, North Carolina. Known 1863. Of. 14 gr. 14 cm. Damascened, violet and blue (Kamacite), pink (Taenite). Homestead, Iowa. Fell February 12, 1875. Cgb. 62 gr. 11cm. Gray unchanged mass (48 gr.), green (serpentinized) mass (14 gr.). Ophir, Montana. Found 1897. Dn. 30 gr. Chips 5-7 em. long, o.5-1 mm. thick. V. WEATHERING, FoRMATION oF NEw CONSTITUENTS. Saline, Kansas. Fell November 15, 1898. Cck. 67 gr. 22cm. Spots of rust piercing through the crust. Stalldalen, Sweden. Fell June 28, 1870. Cgb. 37 gr. 9 cm. Limonite formed on a harness-face. Long Island, Kansas. Found 1892. Cia. 160 gr. 30 cm. Rusted through the whole mass by resting in moist earth. Amana, Somaliland. Fell July 4, 1889. Ckb. 93 gr. 30 em. Stratified Limonite crust on the surface. Amana. 8 gr. 3 cm. Loose stratified Limonite crust. 1904.] BREZINA—COLLECTIONS OF METEORITES. 223 Mackinney, Kansas. Fell 1870. Cs. 91 gr. 40cm. Limonite crust on the surface. Mackiney. 50 gr. Loose Limonite crust. Orgueil, France. Fell May 14,1864. K. 33 gr. Weath- ering grains. Brenham, Kansas. Found 1885. Pk. 98 gr. 29 cm. Olivines partly fresh, partly limonitized. Brenham. 187 gr. 36cm. Olivines browned. Brenham. 75 gr. Loose limonitized grains as products of weathering. Mount Dyrring, New South Wales.” Known 1go2. Pk. , 26 gr. g cm. Olivines browned, Nickel-iron limonitized, Schreibersite fresh. Mount Dyrring. 149 gr. 35 cm. Nearly entirely limoni- tized; weathering in layers; olivines mostly changed into red- dish-white substances. Admire, Kansas. Found 1892. Pr. 244 gr. 20 cm. Weathering-crust 5-10 mm. thick; beginning of falling to pieces by formation of fissures. Imilac, Chile. Found 1800. Pi. 100 gr. Twenty-three weathering individuals. ° Wichita, Texas. Found 1836. Og. 44gr.14cm. Worm- like limonitic rust-figures, free of Bacteria. Joe Wright, Arkansas. Found 1884. Om. 15 gr. 4 cm. Dividing along the octahedron-faces. Lipan Flats, Texas. Found 1897. Om. 184 gr. 27 cm. Dividing partly along octahedron-faces, partly curvilinearly. Tarapaca, Chile. Known 1894. Om. 264 gr. 22 cm. Uncovered (disintegrated) octahedron-faces of 3-4 cm. by weathering on fresh mass. Ranchito, Mexico. Found 1871. Off. 65 gr. 12 cm. Uncovered octahedron-faces of 2 cm. by weathering on fresh mass. Cosby’s Creek, Tennessee. Found 1837. Og. 37 gr.6cm. Uncovered (disintegrated) octahedron with superposed Tae- ' nite lamelle. Welland, Canada. Found 1888. Om. 11 gr. Uncovered octahedron lamelle. 224 BREZINA—COLLECTIONS OF METEORITES, [April 8, Nelson County, Kentucky. Found 1860. Ogg. 261 gr. 40cm. Octahedral weathering on fresh mass. Sao Juliao, Portugal. Found 1883. Ogg. 14 gr. Nail formed by weathering. Mount Joy, Pennsylvania. Found 1887. Ogg. 29 gr. Fallen down in grains of 2-3 cm. Badger, New Mexico. Known 1897. Om. 5 gr. Grains of Limonite as products of weathering. Apolonia, Mexico. Found 1897. 40 gr. 12 cm. Changed into Limonite. Sao Francisco, Brazil. Found 1874. Tell. 343 gr. 28 cm. Penetrating of limonitic alteration in layers. Sao Francisco. 47 gr.8cm. Altered to Hematite. Sao Francisco. 89 gr.1g9cem. Cellular alteration to Hema- tite and Limonite. Augustinowka, Russia. Found 1890. Of. 130 gr. 22 cm. Stratified Limonite-crust of 2-3 cm. thickness. San Cristobal, Chile. Found 1882. Dl. 5 gr. Whitish- limonitic products of alteration. Vaca Muerta, Chile. Known 1861. Mg. 63 gr. 12 cm. Forming of Nickel sulphates. Dofia Inez, Chile. Found 1888. M. 15 gr.6cm. Form- ing of Nickel sulphates. VI. CONSTITUENTS OF METEORITES. Saline, Kansas. Fell November 15, 1898. Cck. 146 gr. 36cm. Free Phosphorus by opening the interior. Nowo Urej, Russia. Fell September 22, 1886. U. 14 gr. 6 cm. Diamond as microscopic component, one per cent. of mass, Carcote, Chile. Known 1888. Ck. Splinters. Diamond as microscopic component. Badger, New Mexico. Known 1897. Om. 537 gr. 42 cm. Graphite with attached Troilite in form of a T, 3 to 4 cm. Plate V, Fig. 45. Toluca, Mexico. Described 1784. Om. 177 gr. 22 cm, Graphite as layer between nuggets of Troilite and their mantle of Schreibersite. 1904.] BREZINA—COLLECTIONS OF METEORITES. 925 Senhadja, Algiers. Fell August 25, 1865. Cwa. 145 gr. zocm. Crystals of Nickel-iron, partly with cleavage faces, in Troilite; chondrule of 11 mm. diameter with bar. Vaca Muerta, Chile. Known 1861. Mg. 1 gr. Grain of Nickel-iron with Widmanstatten-figures. Crab Orchard, Tennessee. Found 1887.-Mg. 35 gr. 12 em. Chondrule of Nickel-iron with Widmanstatten-figures. Mincy, Missouri. Found 1856. M. 140 gr. 30 cm. Chondrule of Nickel-iron, 2 cm. diameter, with worm-like residues of Silicates. Morristown, Tennessee. Found 1887. Mg. ror gr. 35 em. Chondrules of Nickel-iron, 1-3 cm. diameter, with Sili- cate grains. Hainholz, Germany. Found 1856. M. 77 gr. 19 cm. Chondrules of Nickel-iron, 5s-1o mm. diameter. Baratta, New South Wales. Fell May, 1845. Cgb. 58 gr. 18cm. Chondrules partly with Iron cover, partly with Troi- lite cover. Mackinney, Texas. Fell 1870. Cs. 300 gr. 40cm. Black chondrules with Iron cover. Marjalathi, Finland. Fell June 1, 1902. Pi. 61 gr. 17 em. Crystals of Nickel-iron up to 6 mm. diameter, with rounded edges; crystals of Chromite; cylinder of Troilite. Sao Francisco, Brazil. Found 1874. Tell. 24 gr. 6 cm. Crystals of Nickel-iron with folded faces. Ovifac, Greenland. Found 1808. Tell. 242 gr. 50 cm. Nickel-iron grains in Basalt. Ovifac. 225 gr. 42 cm. Chondrules of Nickel-iron in Basalt. Plate VI, Fig. 46. Ovifac. 176 gr. 40 cm. Veins of Nickel-iron in Basalt. Coahuila, Mexico. Known 1837. H. 293 gr. 30 cm. Kamacite with hexahedral cleavage. Cation Diablo, New Mexico. Found 1891. Og. 118 gr. 20cm. Consisting of Kamacite. Toluca, Mexico. Described 1784. Om. 347 gr. 30 cm. Kamacite with hatchings. Merceditas, Chile. Known 1884. Om. 177 gr. 38 cm. Kamacite with hatchings. PROC. AMER. PHILOS, SOC. XLIII. 176.0. PRINTED JUNE 27, 1904. 226 BREZINA—COLLECTIONS OF METEORITES. [April 8, Seelasgen, Germany. Found 1847. Ogg. 52 gr. 10 cm. Kamacite with strong orientated glitter. Pila, Mexico. Known1804. Om. 240gr.30cm. Kama- cite sparkling. Fort Pierre, Missouri. Found 1856. Om. 330 gr. 43 cm. Iron dull. Plymouth, Indiana. Found 1893. Om. 45 gr. 28 cm. Iron dull. Nelson County, Kentucky. Found 1860. Ogg. 103 gr. 38 cm. Iron with silky luster. Walker Township, Michigan. Found 1886. Of. 333 gr. 45cm. Kamacite banded. Burlington, New York. Known 1819. Om. to gr. 3 cm. Kamacite puffy. Brenham, Kansas. Found 1885. Pk. 267 gr. 28 cm. -Wrapping-Kamacite. Welland, Canada. Found 1888. Om. ggr. Taenites iso- lated by weathering. Bella Roca, Mexico. Described 1888. Of. 85 gr. 28 cm. Taenite prevailing. LaCaille, France. Found 1600. Om. 25 gr.14cm. Tae- nite prevailing beneath sparkling Kamacite and Plessite. Misteca, Mexico. Described 1804. Om. 128 gr. 42 cm. Taenite with fernlike skeletons. Thunda, Australia. Described 1886. Om. 49 gr. 9 cm. Taenite strongly developed. Coopertown, Tennessee. Known1860. Om. 49 gr. 11cm, Skeletons of Taenite. Carlton, Texas. Found 1887. Off. 131 gr.33 cm. Ples- site prevailing. Mungindi, Australia. Known 1897. Off. 47 gr. 17 cm. Plessite prevailing, with central skeletons, Thurlow, Canada. Found 1888, Of. 22 gr.6cm. Ples- site with central skeletons. Toluca (doubtful). 27 gr. 6 cm. Bridges (bars) through fields and between puffy beams. Tazewell, Tennessee. Found 1853. Off. 100 gr. 20 cm. Dodecahedral lamella beneath octahedral ones, ——————— eC 1904.] BREZINA—COLLECTIONS OF METEORITES. 227 Bel'a Roca, Mexico. Known 1888. Of. 37 gr. 12 cm. Iron-tongue in Troilite. San Cristobal, Chile. Found 1882. Db. 3or gr. 35 cm Shagreen Iron in meandering latticed Iron. San Cristobal. 29 gr. 2 cm. Gold-yellow crystal in Troi- lite. Beaconsfield, Australia. Found 1897. Og. 1 gr. Iso- lated crystals of Cohenite. Niakornak, Greenland. Found 1819. Tell. 1 gr. Iso- lated Cohenite crystals. Glowed steel with 1.3 percent. C. Artificial. 1 gr. Iso- lated Cohenite crystals. Ruffs Mountain, South Carolina. Known 1850. Om. 47 gr.12cm. Ribs of gray Cohenite in Kamacite. Bendego, Brazil. Found 1784. Og. 148 gr. 25 cm. Cohenite-ribs in Kamacite, porous. Wichita, Texas. Found 1836. Og. 428 gr. 40 cm. Cohenite ribs with high lustre in Kamacite. Magura, Hungary. Found 1840. Og. 174 gr. 36 cm. Cohenite ribs in Kamacite, united to skeletons. Penkarring Rock, Australia. Found 1864. Og. 73 gr- 11cm. Cohenite ribs and Schreibersite in Kamacite. Cation Diablo, New Mexico. Found 1891.. Og. 159 gr. 23cm. Cohenite ribs in Kamacite porous and compact. Rosario, Honduras. Known 1897. Og. 18 gr. 7 cm. Cohenite lamellz and skeletons in Kamacite. Deep Springs Farm, North Carolina. Found 1846. Db. 30 gr. 5 cm. Orientated plates of crystals of Cohenite in dull Iron. Ovifac, Greenland. Found 1808. Tell. 152 gr. 42 cm. Cohenite forming with Nickel-iron grains in Basalt. Sao Juliao, Portugal. Found 1883. Ogg. 2 gr. Isolated crystals of Schreibersite, iridescent. Sao Juliao. roo gr. Isolated crystals of Schreibersite. Sao Juliao. 60 gr.12 cm. Uncovered skeleton of Schrei- bersite. Carlton, Texas. Found 1887. Off. 5 gr. Isolated Schrei- bersite crystals. - 228 BREZINA—COLLECTIONS OF METEORITES. [April 8, Carlton. 277 gr.31cm. Crystals of Schreibersite in Wrap- ping Kamacite in the Trias. Primitiva, Chile. Found 1888. Dp. 4gr. Isolated frag- ments of Schreibersite crystals. Toluca, Mexico. Described 1784. Om. 119 gr. 13 cm. Crystals of Schreibersite with smooth faces and rounded edges; Graphite. Saint Francois, Missouri. Known 1863. Og. 47 gr. 30 em. Schreibersite ribs in Kamacite. Bischtttbe, Russia. Found 1888. Og. 263 gr. 40 cm. Skeleton-like crystals of porous Schreibersite in parallel layers of compact Schreibersite in wrapping-Kamacite in the Trias. Tennant’s Iron. Found 1784. Og. 9 gr. 4 cm. Crystals of Schreibersite in parallel layers of Cohenite beneath free Cohenite crystals. Dacotah, Indian Territory. Found 1863. Ogg. 90 gr. 35 cm. Hieroglyphic Schreibersite, partly faulted by a fis- sure. DeSotoville, Alabama. Found 1878. H. 298 gr. 45 cm. Schreibersite crystals partly turning into Schreibersite hiero- glyphs. DeSotoville. 287 gr. 40 cm. Crystal of Schreibersite in Limonite-Magnetite beneath Schreibersite hieroglyphs and Rhabdite ranges. San Cristobal, Chile. Found 1882. Dl. 14 gr. 7 cm. . Two layers of Schreibersite (compact inward, grainy outward) on Troilite. San Cristobal. 67 gr.12cm. Schreibersite with hatchings in a Troilite crystal. Ballinoo, Australia. Found 1893. Off. 99 gr. 28 cm. Schreibersite points in Troilite nuggets. Ballinoo. 8 gr.8cm. Three loose Troilite nuggets, partly with crystalline surface, with points of Schreibersite. Sao Juliao, Portugal. Found 1883. Ogg. 280 gr. 40 cm. Plates of Giant-Rhabdites, 2 cm. long, terminating hiero- glyphs of Schreibersite. Locust Grove, North Carolina, Found 1857. Ds. 206 gr. 38cm. Rhabdite plates of 5-12 mm., apparently orientated. Seelasgen, Germany. Found 1847. Ogg. 20 gr. 5 cm. 1904. | BREZINA—COLLECTIONS OF METEORITES. 929 Rhabdites abundant in Kamacite; etching zones round Taenites, Hex River Mounts, Capeland. Found 1882. H. 216 gr. 45 cm. Parallel ranges of diagonal Rhabdites. DeSotoville, Alabama. Found 1878. H. 266 gr. 43 cm. Parallel ranges of diagonal Rhabdites beneath skeletons of Schreibersite. Plate VI, Fig. 47. Scottsville, Kentucky. Found 1867. H. 57 gr. 14 cm. Wreath of Rhabdites around Troilite-grain. Fort Duncan, Texas. Found 1882. H. 99 gr. 24 cm. Inversion of orientated glitter inward and outward of etching zone of Troilites; Neumann-lines traversing; rust figures worm-like. Flovd Mountain, Virginia. Found 1887. Ha. 400 gr. 40cm. Spot-ranges and Rhabdite-ranges in parallel planes; Troilite with Cohenite or Schreibersite points. Butler, Missouri. Found 1874. Off. sgr. Isolated Troi- lites limonitized. Sao Francisco, Brazil. Found 1874. Tell. 2 gr. 1 cm. Troilite crystals on folded Troilite plates. Kansada, Kansas. Found 1894. Cib. 194 gr. 25 cm. Troilite nuggets of various forms beneath Nickel-iron grains in Chondrite. Zavid, Bosnia. Fell August 1, 1897 Cia. 67 gr. 14 cm. Nest of 3 cm. diameter of Troilite grains in Chondrite. MacKinney, Texas. Fell 1870. Cs. 250 gr. 40 cm. Troilite vein 3-10 mm. thick. MacKinney. 351 gr.48cm. Troilite grains of 3-7 mm. in Chondrite. Baratta, New South Wales. Fell May, 1845. Cgb. 138 gr. 38 cm. Chondrules with Troilite mantle beneath grains of Troilite and Nickel-iron. Bella Roca, Mexico. Known 1888. Of. 94 gr. 30 cm. Nuggets of Troilite with wrapping-Kamacite in the Trias. Mukerop (Bethany), Namaland. Found 1899. (1853.) Of. 363 gr. 41cm. Nuggets of two kinds of Troilite; the soluble, with fluidal structure interlocking as tongues in the insolu- ble one. Mukerop. 489 gr. 45 cm. Nuggets of soluble and insolu- 230 BREZINA—COLLECTIONS OF METEORITES. [April 8, ble Troilite; Reichenbach lamelle consisting of bulbous Troi- lite. Plate VII, Fig. 48. Lagrange, Kentucky. Found 1860. Of. 29 gr. 18 cm. Reichenbach lamelle with bent borders, trailed. Merceditas, Chile. Known 1884. Om. 115 gr. 41 cm. Reichenbach lamelle originating in Troilite nuggets. Joe Wright, Arkansas. Found 1884. Om. 135 gr. 40cm. Reichenbach and Schreibersite lamelle 2-3 cm. long. Trenton, Wisconsin. Found 1858. Om. 57 gr. 30 cm. Reichenbach lamelle 2-3 cm. long. Primitiva, Chile. Found 1888. Dp. 25 gr.14cm. Troi- lite nuggets with Iron tongues in Nickel-iron, small Troilite globes swarming around, beneath hieroglyphs of Schreibers- ite. Sao Julia, Portugal. Found 1883. Ogg. 157 gr. 20 cm. Troilite points in parallel ranges beneath crystals and hiero- glyphs of Schreibersite. Bendego, Brazil. Found1784. Og. 54gr.13cm. Crys- tal of Daubréelite with adhering Troilite, 1 cm. diameter. Santo Domingo Yanhuitlan, Mexico. Known 1804. Of. 42 gr. 36cm. Oval Troilite nuggets transversed by bands of Daubréelite; Reichenbach lamelle. Badger, New Mexico. Known1897. Om. 434 gr. 33 cm. Daubréelite crystal in Troilite nugget, the whole in wrapping Kamacite, on which stowed the octahedral lamelle; half detached octahedron. Shalka, India. Fell November 30, 1850. Chl.. 10 gr. 4cm, Chromite individuals up to 5 mm. diameter, strongly deformed. Marjalathi, Finland. Fell June 1, 1902. Pi. 192 gr. 30 em. Crystals of Chromite, uncovered and in section. Mount Dyrring, New South Wales. Known 1902. Pk. 52 gr. 20 cm. Crystals of Chromite, octahedron, dodeca- hedron, trapezohedron and two hexakisoctahedra. Krasnojarsk, Siberia. Found 1749. Pk. 4 gr. Isolated crystals of olivine, olive-green and brown, Jamyschewa, Siberia. Found 1885. Pk. 2gr. Isolated olivine crystals. 1904.] BREZINA—COLLECTIONS OF METEORITES. 231 Brenham, Kansas. Found 1885. Pk. 50 gr. Isolated olivine crystals. Brenham. 64 gr. 33cm. Parallel laces of olivine crystals in Nickel-iron. Estherville, Iowa. Fell May 10, 1879. M. 20 gr. Iso- lated individual of olivine. Mincy, Missouri. Found 1856. M. 154 gr. 42cm. Oli- vine crystal 5 cm. diameter, in Mesosiderite. Eagle Station, Kentucky. Found 1880. Pr. 86 gr. 25 cm. Olivine crystals up to 2 cm. diameter, broken, with inserted Nickel-iron between the fragments. Vaca Muerta, Chile. Found 1861. Mg. Three thin sec- tions of an olivine crystal. Stannern, Moravia. Fell May 22, 1808. Eu. 130 gr. 25 em. Vein of Anorthite in normal Eukrite. New Concord, Ohio. Fell March 1, 1860, Cia. 74 gr. 15cm. Grains of Anorthite in Chondrite. Toluca, Mexico. Found 1784. Om. Microscopic prepa- ration of crystals of Kosmochlore. Saint Caprais, France. Fell January 28, 1883. Ci. 1 gr. Icm. Greenish-yellow crystals of Enstatite. Fisher, Minnesota. Fell April 9, 1894. Cia. 10 gr. 6 cm. Foliated chondrule of Enstatite, 1 cm. diameter. Hvittis, Finland. Fell October 21, 1rg0r1. Cek. 11 gr. 3.cm. Enstatite chondrule, 1 cm. diameter. Alfianello, Italy. Fell February 16, 1883. Ci. 36 gr. 12 em. Chondrules black, gray and striated (black and white). Zavid, Bosnia. Fell August 1, 1897, Cia. 24 gr. 7 cm. Greenish-gray, fragmentary chondrule of 1 cm. diameter in Chondrite. Zavid. 71 gr. 12 cm. Dark gray chondrule with white, cruciform skeleton. Antifona, Italy. Fell February 3, 1890, Cc. 2 gr. 3 cm. White chondrule with blue nucleus. Kaande, Russia. Fell May 11, 1855. Cw. 12 gr. 4 cm. White chondrule of 5 mm. diameter. _ Bath, Dakota. Fell August 29,1892. Ccb. 18 gr. 5 cm. Radiated chondrule of white and dark gray sectors with black mantle. 232 BREZINA—COLLECTIONS OF METEORITES, [April 8, Chantonay, France. Fell August 5,1812. Cgb. 176 gr. 35 cm. Chondrule of 1 cm. diameter, reticulated hexagonally. Bjurbéle, Finland. Fell March 12, 1899. Cca. 128 gr. 20 cm. Gray, oval chondrule of 1 to 1.3 cm. diameter; iri- descent Troilite. Bjurbéle. 5 gr. Isolated chondrules. Allegan, Michigan. Fell July 10, 1899. Cco. 7 gr. Iso- lated chondrules with drusy surface. MacKinney, Texas. Fell 1870. Cs. 301 gr. 38 cm. Olive-green, black and Troilite-bearing chondrules up to 1 cm. diameter. MacKinney. 280gr. 40cm. Leek-green cross-chondrules, 1 cm. diameter. MacKinney. 260gr.38cm. Dull black, rectangular crys- talline inclusion. MacKinney. 300gr.42cm, Yellow and glassy chondrules. Baratta, New South Wales. Fell May, 1845. Cgb. 214 gr. 40cm. Chondrule with black and white faulted halves; black glassy chondrules partly with Nickel-iron, partly with Troi- lite mantle. Baratta. 196 gr. 42 cm. White, gray, yellow chondrules and black glassy chondrules. Baratta. 159 gr.39cm. Radiated Troilite-bearing chon- drule with Iron mantle; fragmentary chondrules. Manbhoom, India. Fell December 22, 1863. Am. 3 gr. 2cm. Nugget of crystalline Chondrite Ck, isolated from Am- photerite. Netschajevo, Russia. Found +1846. Obn. 20 gr. § cm. Nugget of veined crystalline Chondrite Cka, isolated from Octahedrite. Devitrified molten pitchstone in the form of a radiated globe with adhering fragments of glass; 6 cm. diameter. VII. System or METeEoRITES. ' STONES—SILICATES PREVAILING. A. Acnonprites.—Stones poor in Nickel-iron, essentially without round chondrules. ' All groups are defined, whether represented in the collection or not. The weights and Roman numbers in parentheses, e. g., Shalka (10 gr. VI), refer to specimens listed in foregoing sections (VI constitu. ents, etc.). 1904.] BREZINA—COLLECTIONS OF METEORITES. 233 1. Chladmite, Chl. Chiefly Bronzite. Shalka, India. Fell November 30, 1850. (10 gr., VI.) Ibbenbithren, Germany. Fell June 17, 1870. 4 gr. 3 cm. 2. Veined Chladuite, Chla. Bronzite with black or metallic veins. 3. Angrite, A: Chiefly Augite. 4. Chassignite, Cha. Chiefly Olivine. Chassigny, France. Fell October 3, 1815, Spl. 1 cm. 5. Bustite, Bu. Bronzite with Augite. 6. Amphoterite, Am. Bronzite with Olivine. Manbhoom, India. Fell December 22, 1863. 219 gr. 30 cm. Three sections. (3 gr., VI.) ? Jelica, Servia. Fell December 1, 1889. 126 gr. 25 cm. Seven sections. 7. Rodite, Ro. Bronzite with Olivine, breccia-like. Bandong, Java. Fell December 10, 1871. 46 gr. 9 cm. 8. Eukrite, Eu. Augite with Anorthite. Constantinople, Turkey. Fell June, 1805. Spl. Stannern, Moravia. Fell June 22, 1808. Two sections. (162 gr., IV; 130 gr., VI.) Juvinas, France. Fell June 1s, 1821. (21 gr., IV.) 9. Shergottite, She. Augite with Maskelynite. 10: Howardite, Ho. Bronzite, Olivine, Augite and Anor- thite. La Vivionnére, France. Fell July 14, 1845. Spl. Zmen, Russia. Fell August, 1858. Spl. 11. Breccia-like Howardite, Hob. Bronzite, Olivine, Augite and Anorthite, breccia-like. ; 12. Leucituranolite, L. Leucite, Anorthite, Augite and glass. B.—CuHoNDRITES.—Bronzite, Olivine and Nickel-iron, with round or round and polyhedric chondrules. 13. Howarditic Chondrite, Cho. Polyhedric secretions pre- vailing, round chondrules scarce. Crust partly bright. Borgo San Donino, Italy. Fell April 19, 1808. Two thin sections, Krahenberg, Germany. Fell April 5, 1869. 2 gr. 1 cm. Ottawa, Kansas. Fell April 9, 1876. 65 gr. 12cm. One section. 234 BREZINA—COLLECTIONS OF METEORITES. [April 8, 14. Vetned Howarditic Chondrite, Choa. Polyhedric secre- tions prevailing, round chondrules scarce. Black or metallic veins. £5. White Chondrite, Cw. White, rather friable mass with scarce, mostly white chondrules. Mordvinovka, Russia. Prehistoric. (1 gr., II.) Mauerkirchen, Upper Austria. Fell November 20, 1768. 3 gr. 2 cm. . Linum, Germany. Fell September 5, 1854. Spl. 1 cm. Kaande, Livland. Fell May 11, 1855. 7 gr. 4 cm. (12 gr., VI.) Tourinnes, Belgium. Fell December 7, 1863. 16 gr. 5 cm. San Pedro Springs, Texas. Found 1887. 4gr.2cm. Pricetown, Ohio. Fell February 13, 1893. 2 gr. 1 cm. 16. Veined White Chondrite, Cwa. White, rather friable mass with scarce, mostly white chondrules; black or metallic veins. Lucé, France. Fell September 13, 1768. Two thin sec- tions. Wold Cottage, England. Fell December 13, 1795. Spl. Kuleschowka, Russia. Fell March 12, r81r. 1 gr. r cm. Honolulu, Hawaii. Fell November 27, 1825. 6 gr. 3 cm. Drake Creek, Tennessee. Fell May 9, 1827. 2 gr. 1 cm. Aumiéres, France. Fell June 3, 1842. (36 gr., IV.) Schénenberg, Bavaria. Fell December 25, 1896. 7 gr. 2 cm. Marion, lowa. Fell February 25, 1847. 40 gr. 9 cm. Girgenti, Italy. Fell February 10, 1853. (132 gr., IV.) Scheikahr, Curland. Fell June 2, 1863. 66 gr. 30 cm. Senhadja, Algiers. Fell August 25, 1865. (145 gr., VI.) Grossliebenthal, Russia. Fell November 19, 1881. 79 gr. 16 cm. Mocs, Hungary. Fell February 3, 1882. Thirty thin sec- tions. (141 gr., IV.) 17. Breccia-like White Chondrite, Cwb. White, rather fri- able mass with scarce, mostly white chondrules; breccia-like. Lissa, Bohemia, Fell September 3, 1808. 3 gr. 3 cm. Aleppo, Turkey. Found 1873. (67 gr., IV.) Pacula, Mexico, Fell June 18, 1881. 2 gr. 3 cm, 1904.] BREZINA—COLLECTIONS OF METEORITES. 235 18. Intermediate Chondrite, Ci. Firm, polishable mass with white and gray chondrules breaking with matrix Dhurmsala, India. Fell July 14, 1860. (257 gr., IV.) Rakowka, Russia. Fell November 20, 1878. 130 gr. 30 cm. . Saint Caprais, France. Fell January 28, 1883. (1 gr., VI.) Alfianello, Italy. Fell February 16, 1883. 286 gr. 38 cm. (36 gr., VI.) 19. Veined Intermediate Chondrite, Cia, Firm, polishable mass with white and gray chondrules breaking with the matrix; black or metallic veins. Berlanguillas, Spain. Fell July 8, 181rz. 1 gr. 1 cm. Durala, India. Fell February 18, 1815. 37 gr. 8 cm. Vouillé, France. Fell May 13,1831. 31 gr. 7 cm. Macao, Brazil. Fell November 11, 1836. 3 gr. 2 cm. Chateau-Renard, France. Fell June 12, 1841. 307 gr. 33cm. (80 gr., IV.) Mainz, Germany. Found 1852. 26 gr. 6 cm. New Concord, Ohio. Fell May 1, 1860. (74 gr., VI.) Nerft, Curland. Fell April 12, 1864. 98 gr. 20 cm. Maémé, Japan. Fell November 10, 1886. (97 gr., IV.) Long Island, Kansas. Found 1892. 173 gr. 35 cm. {160 gr., V.) Zabrodje, Russia. Fell September 22, 1893. 4 gr. 3 cm. Fisher, Minnesota. Fell Aprilg, 1894. (208 gr., IV; rogr., VI.) Bori, India. Fell May 9, 1894. 12 gr. 5 cm. Lancon, France. Fell June 20, 1897. 1 gr. 1 cm. Zavid, Bosnia. Fell August 1,1897. 147 gr.36cm. Four sections. (310 gr., IV; 162 gr., VI.) Gambat, India. Fell September 15,1897. 1 gr. 1 cm. Bath Furnace, Kentucky. Fell November 15, 1902. 5 gr. 3cm. 20. Breccialike Intermediate Chondrite, Cib. Firm, polish- able mass, white and gray chondrules breaking with matrix; breccialike. ; Laigle, France. Fell April 26, 1803. 47 gr.12cm. Four sections. (115 gr., II.) Saint Mesmin, France. Fell May 30, 1866. 8 gr. 4 cm. 236 BREZINA—COLLECTIONS OF METEORITES. {April 8, Laborel, France. Fell June 14,1871. 3 gr. 2cm. Bjelokrynitschie, Russia. Fell January 1, 1887. 7 gr. 5 cm. Kansada, Kansas. Found 1894. 25 gr.g cm. (135 gr., IV;, 194 gr., VI.) 21. Gray Chondrite, Cg. Firm, gray mass, chondrules of various kinds breaking with matrix. Knyahinya, Hungary. Fell June 9, 1866. Eight thin sec- tions. (423 gr., IV.) 22. Veined Gray Chondrite, Cga. Firm, gray mass, chon- drules of various kinds breaking with matrix; black or metal- lic veins. Barbotan, France. Fell July 24, 1890. (9 gr., Il.) Charsonville, France. Fell November 23, 1810. 153 gr. 22cm. Lasdany, Russia. Fell July 12, 1820. (77 gr., IV.) Parnallee, India. Fell February 28,1857. 40 gr. 11 cm. Alessandria, Italy. Fell February 2, 1860. 137 gr. 20cm. (15 gr., IV.) Lerici, Italy. Fell January 30, 1868. (6 gr., III.) Kerilis, France. Fell November 26, 1874. 39 gr., 14 cm. Cronstadt, Orange River Free State. Fell November 19, 1877. Spl. 41cm. 23. Breccialike Gray Chondrite, Cgb. Firm, gray mass, chondrules of various kinds breaking with matrix; breccia- like. Chantonnay, France. Fell August 5, 1812. One thin sec- tion. (89 gr., IV; 176 gr., VI.) Borodino, Russia. Fell September 5-6, 1812. (5 gr., II.) Baratta, New South Wales. Fell May, 1845. (45'gr., IV; 765 gr., VI.) Mez®-Madarasz, Hungary. Fell September 4, 1852. 67 gr. 14cm. Elgueras, Spain. Fell December 6, 1866. 16 gr. 6 cm. Pultusk, Russia. Fell January 30, 1868. Four sections. (57 gr., III; 639 gr., IV.) Homestead, Iowa. Fell February 12, 1875. (62 gr., IV.) Stalldalen, Sweden, Fell June 28, 1876. (34 gr., IV; 37 gr., V.) Midt Vaage, Norway. Fell May 20, 1884. 1 gr., 1 cm. 1904.] BREZINA—COLLECTIONS OF METEORITES. 237 24. Orvinite, Co. Black infiltrated mass, fluidal texture; surface uneven, crust interrupted. \ Orvinio, Italy. Fell August 31, 1872. 1 gr.1cm. One section. (30 gr., IV.) 25. Tadjerite, Ct. Black, half-glassy, crust-like mass with- out crust on surface. 26. Black Chondrite, Cs. Dark or black mass, chondrules of various kinds breaking with matrix. Mikenskoi, Russia. Fell June 28, 1861. 2gr.2cem. Two sections. MacKinney, Kansas. Fell 1870. 243 gr. 45 cm. Five sections. (141 gr., V; 2002 gr., VI.) Sevrukof, Russia. Fell May 11, 1874. 19 gr. 8 cm. Tschuwaschskaja, Russia. Found 1898. 15 gr. 7 cm. 27. Veined Black Chondrite, Csa. Dark or black mass, chondrules of various kinds, breaking with matrix; black or metallic veins. Farmington, Texas. Fell June 25, 1890. 68 gr. 14 cm. Two sections. 28. Ureilite, U. Black mass, chondritic or granular; Iron in veins or incoherent. Nowo Urej, Russia. Fell September 22, 1886. (14 gr., VI.) ; 29. Coaly Chondrite, K. Dull black, friable chondrite with free carbon, low specific gravity, metallic Iron nearly or wholly wanting. Cold Bokkeveld, Cape Colony. Fell October 13, 1838. 2gr.1cm. One section. Orgueil, France. Fell May 14,1864. 6gr.5cm. (33 gr., V.) Nogoya, Argentina. Fell July 1, 1879. 1 gr., 1 cm. Mighei, Russia. Fell June 18,1889. 12 gr. 4 cm. 30. Globular Coaly Chondrite, Ke. Dull gray or black, friable mass with free carbon; chondrules not breaking with matrix; metallic Nickel-iron. 31. Veined Globular Coaly Chondrite, Kca. Dull black firm mass with free carbon; chondrules not breaking with matrix; metallic veins. Indarch, Russia. Fell April7, 1891. 136 gr. 16 cm. 238 BREZINA—COLLECTIONS OF METEORITES. [April 8, 32. Globular Chondrite, Cc. Friable mass with hard (radi- ated) chondrules not breaking with matrix. Albareto, Italy. Fell July 1766. (3 gr., II; Spl. IV.) La Baffe, France. Fell September 13, 1822. 2 gr. 1 cm. Praskoles, Bohemia. Fell October 14, 1824. 11 gr. 5 cm. Le Pressoir, France. Fell January 25,1845. 10 gr. 4 cm. Yatoor, India. Fell January 23, 1852. 77 gr. 16 cm. Avilez, Mexico. Fell June, 1856. 5 gr. 2 cm. Quenggouk, India. Fell December 27,1857. 65 gr. 12cm. Aussun, France. Fell December 9, 1858. 110 gr. 17 cm. Motta di Conti, Italy. Fell February 29, 1868. 33 gr. 10 cm. Hessle, Sweden. Fell January 1, 1869. 132 gr. 24 cm. (23 gr., IV.) Sarbanovac, Servia. Fell October 13, 1877. 1 gr. Two sections. Tieschitz, Bohemia. Fell July 15, 1878. 3 gr. 1 cm. Three sections. Gnadenfrei, Germany. Fell May 17, 1879. 1 gr. 1 cm. Torre, Italy. Fell May 24, 1886. 3 gr. 2 cm. Antifona, Italy. Fell February 3, 1890. 232 gr. 33 cm. (241 gr., IV; 2 gr., VI.) Misshof, Curland. Fell April 10, 1890. 69 gr. 11 cm, Mount Browne, New South Wales. Fell July 17, 1902. 81 gr. 25 cm. 33. Vetned Globular Chondrite, Cca. Friable mass with hard (radiated) chondrules not breaking with matrix; black or metallic veins. Trenzano, Italy. Fell November 12, 1856. 171 gr. 22 cm. Bjurbdle, Finland, Fell March 12, 1899. 199 gr. 32 cm. (61 gr., II; 133 gr., VI.) 34. Breccia-like Globular Chondrite, Ccb. Friable, breccia- like mass with hard (radiated) chondrules not breaking with matrix, Krawin, Bohemia, Fell July 3, 1753. 6 gr. 3 cm. Weston, Connecticut. Fell December 14, 1807. 23 gr. 4 cm. Mooresfort, Ireland. Fell August, 1810, 14 gr. 6 cm. Cereseto, Italy. Fell July 17, 1840. 1 gr. 1 cm. 1901.) BREZINA—COLLECTIONS OF METEORITES. 239 / Kesen, Japan. Fell May 13, 1850. (8 gr., II.) Gnarrenburg, Germany. Fell May 13, 1855. Spl. Waconda, Kansas. Found 1874. 35 gr. 11 cm. Ochansk, Russia. Fell August 30, 1887. 95 gr. 16 cm., 16 gr.5cm. (12 gr., IV.) Forest, Iowa. Fell May 2, 1890. Nine sections (83 gr., t¥) Bath, Dakota. Fell August 29, 1892. 300 gr. 40 cm. (18 gr., VI.) 35. Ornansite,Cco. Friable mass of chondrules. Ornans, France. Fell July 11, 1868. 1 gr. 1 cm. Warrenton, Missouri. Fell January 3, 1877. 15 gr. 7 cm Allegan, Michigan. Fell July 10, 1899. 232 gr. 35 cm. (7 gr., VI.) 36. Ngawite, Cen. Friable breccialike mass of chondrules. 37. Crystalline Globular Chondrite, Cck. Hardly friable, crystalline mass with hard (radiated) chondrules, partly break- ing with matrix, partly not. Menow, Germany. Fell October 7, 1862. 7 gr. 4 cm. Prairie Dog, Kansas. Found 1893. 96 gr. 16 cm. Beaver Creek, British Columbia. Fell May 26, 1893. 36 gr. 8 cm. Sawtschenskoje, Russia. Fell July 27, 1894. 22 gr. 9 cm. Ambapur, India. Fell May 27, 1895. 22 gr.g cm. Two sections. Saline, Kansas. Fell September 15, 1898. Four sections. (67 gr., V; 146 gr., VI.) Chervettaz, Switzerland. Fell November 30, 1901. 32 gr. 8 cm. 38. Veined Crystalline Globular Chondrite, Ccka. Hardly friable, crystalline, veined mass with hard (radiated) chon- drules partly breaking with matrix, partly not. 39. Breccia-like Crystalline Globular Chondrite, Cckb. Hardly _ friable, crystalline, breccia-like mass with hard (radiated) chondrules partly breaking with matrix, partly not. 40. Crystalline Chondrite, Ck. Hard crystalline mass with hard (radiated) chondrules breaking with matrix. _Pillistfer, Livland. Fell August 8, 1863. 187 gr. 33 cm. Two sections. 240 BREZINA—COLLECTIONS OF METEORITES. [April 8, Tjabé, Java. Fell September 19, 1869. Two sections. Alastoewa, Java. Fell March 19, 1884. 1 gr. 2 cm. Carcote, Chile. Found 1888. (Spl. VI.) Gilgoin, New South Wales. Described 1889. 7 gr. 3 cm. Guarefia, Spain. Fell July 20, 1892. 5 gr. 3 cm. Oakley, Kansas. Found 1895. 195 gr. 38 cm. 41. Veined Crystalline Chondrite, Cka. Hard crystalline veined mass with hard (radiated) chondrules breaking with matrix. Kernouvé, France. Fell May 22, 1869. 122 gr 28 cin. (173 gr., IV.) Pipe Creek, Texas. Found 1887. 87 gr. 31 cm. 42. Breccialike Crystalline Chondrite, Ckb. Hard, crystal- line breccialike mass with hard (radiated) chondrules breaking with matrix. Ensisheim, Germany. Fell November 16, 1492. (10 gr., II.) Bluff, Texas. Found 1878. 258 gr.42cm. Four sections. Amana, Somaliland. Fell July 4, 1889. 174 gr. 36 cm. (101 gr., V.) C. ENSTATITE-ANORTHITE-CHONDRITE. Enstatite, Anorth- ite and Nickel-iron with round chondrules. 43. Crystalline Enstatite-anorthite Chondrite, Cek. Hard crystalline mass with hard (radiated) chondrules. Hvittis, Finland. Fell October 21, 1g01. 87 gr. 32 cm. (11 gr., VI.) D. SrperRoLite.—Transitions of stones to irons. Nickel- iron in the mass cohering, on sections separated. 44. Mesosiderite, M. Crystalline Olivine and Bronzite, Hainholz, Germany. Found 1856. (77 gr., VI.) Mincy, Missouri. Found 1856. 100 gr. 23 cm. (294 gr. VL.) Estherville, Iowa. Fell May 10, 1879. tar gr. 15 cme (40 gr., IV; 20 gr., VI.) Karand, Persia. Fell May, 1880. 26 gr. 10 cm. Inca, Chile. Known 1888. 41 gr. 8 cm. Dofia Inez, Chile. Known 1888. 54 gr. 12 cm. Four sections. (15 gr., V.) > rer ar © 1904.] BREZINA—COLLECTIONS OF METEORITES. 241 45. Grahamite, Mg. Crystalline Olivine, Bronzite and Plagioclase. Vaca Muerta, Chile. Known 1861. 46gr.14cm. 12 sec- tions. (184 gr., III; 63 gr., V; 4 gr., VI.) Crab Orchard, Tennessee. Found 1887. 88 gr. 28 cm. (35 gr., VI.) sa Morristown, Tennessee. Found 1887. (ror gr., VI.) 46. Lodranite,Lo. Granular-crystalline Olivine and Bronz- ite. | IRON-METEORITES. METALLIC CONSTITUENTS PREVAILING OR ALONE. E. LitnHosiperite.—tTransition from stones to irons; Nickel-iron cohering in mass and on sections. 47. Siderophyre, Si. Grains of Bronzite with accessory Asmanite in the Trias. Rittersgrtin (Steinbach), Saxony. Found 1843 (1724). 30 gr. 12 cm. 48. Pallasite-Krasnojarskgroup, Pk. Rounded crystals of Olivine in the Trias. Krasnojarsk, Siberia. Found 1749. (25 gr., II; 4gr., VI.) Mount Vernon, Kentucky. Found 1868. 300 gr., 42 cm. Glorietta, New Mexico. Found 1884. (119 gr., III; 492 gr., IV.) Brenham, Kansas. Found 1885. (359 gr., III; 360 gr., V; 350 gr., VI.) Jamyschewa, Siberia. Found 1885. togr.4cm. (2 gr., VI.) Finmarken, Norway. Found 1902. 250 gr., 40 cm. Mount Dyrring, New South Wales. Known 1go02. (175 gr., V; 52 gr., VI.) 49. Pallasite-Rokitky group, Pr. Polyhedral crystals of Oli- vine partly broken, and fragments separated by Nickel-iron. Eagle, Kentucky. Found 1880. (86 gr., VI.) Admire, Kansas. Found 1892. 74 gr. 22 cm. (244 gr., V.) 50. Pallasite-Imilac group, Pi. Olivine crystals cracked and squeezed. Imilac, Chile. Found 1800. 8g gr.,12 cm. (100 gr. V.) PROC. AMER. PHILOS. SOC. XLIII. 176. P. PRINTED JULY 14, 1904. 242 BREZINA—COLLECTIONS OF METEORITES. [April 8, Marjalathi, Finland. Fell June 1, 1902. 259 gr. 33 cm. (147 gr., IV; 253 gr., VI.) 51. Pallasite-Albach group, Pa. Olivine crystals in fine brec- ciated Trias. F. OcTAHEDRITE.—Kamacite, Taenite and Plessite (Trias), in lamelle and concamerations of the four octahedron faces. 52. Finest Octahedrite, Off. Lamelle up to o.2 mm. thick- ness. Fields prevailing on lamelle. Tazewell, Tennessee. Found 1853. (100 gr., VI.) Ranchito, Mexico. Found 1871. (65 gr., V.) Butler, Missouri. Found 1874. (5 gr., VI.) Carlton, Texas. Found 1887. (147 gr., IV; 413 gr., VI.) Ballinoo, Australia. Found 1893. (395 gr., IV; 107 gr., VI.) Mungindi, New South Wales. Known 1897. (47 gr., VI.) 53. Fine Octahedrite Victoria group, Ofv. Lamelle of Troi- lite and Schreibersite in fine Trias. 54. Fine Octahedrite, Of. Thickness of lamelle o.2-0.4 mm. Santo Domingo Yanhuitlan (Teposcolula), Mexico. Known 1804. (42 gr., VI.) Putnam, Georgia. Found 1839. 22 gr., 5 cm. Bethany (Mukerop, Lion River), Namaland. Found 1853. (1204 gr., III; 1927 gr., IV; 872 gr., VI.) Jewell Hill, North Carolina. Known 1854. 14 gr. 4 cm. Lagrange, Kentucky. Found 1860. (29 gr., VI.) Smith Mountain, North Carolina. Known 1863. (14 gr., IV.) Bickeberg, Germany. Found 1863. 12 gr. 3 cm. : Walker Township, Michigan. Found 1883. (333 gr., VI.) Jamestown, Dacotah. Found 1885. (69 gr., IV.) Bella Roca, Mexico. Known 1888. 82 gr. 28 cm. (104 gr., [V; 216 gr., VI.) Saint Genevieve, Missouri. Found 1888. 313 gr. 45 cm. Thurlow, Canada. Found 1888. (22 gr., VI.) Cuernavaca, Mexico, Described 1889. 5 gr. 4 cm. Apoala, Mexico. Found 1890. 5 gr. 1 cm. Augustinowka, Russia. Found 1890. 39 gr.g cm. (130 gr., V.) a 1904.] BREZINA—COLLECTIONS OF METEORITES. 243 55. Mollified Fine Octahedrite, Ofe. Figures fallen in dis- order by mollifying; points instead of Troilite lamelle. 56. Medium Octahedrite, Om. Thickness of lamelle o.5—1 mm. Casas Grandes, Mexico. Prehistoric. (102 gr., II.) Elbogen, Bohemia. Known 1400. (12@r., II.) LaCaille, France. Known 1600. (25 gr., VI.) Adargas, Mexico. Known 1780. 135 gr. 24 cm. Descubridora, Mexico. Known 1783. 88 gr. 20 cm. Toluca, Mexico.’ Described 1784. (660 gr., VI.) Misteca, Mexico. Described 1784. (128 gr., VI.) Pila, Mexico. Known 1784. (240 gr., VI.) Burlington, New York. Known 1819. (10 gr., VI.) Carthage, Tennessee. Found 1840. 12 gr. 3 cm. Ruffs Mountain, South Carolina. Described 1850. (47 gr., VI.) Fort Pierre, Nebraska. Found 1856. (330 gr., VI.) Staunton IV, Virginia. Found 1858. 51 gr. 18 cm. Trenton, Wisconsin. Found 1858. (57 gr., VI.) Coopertown, Tennessee. Known 1860. (49 gr., VI.) Nejed, Arabia. Found 1864. 25 gr. 11 cm. Caperr, Patagonia. Known 1869. 6 gr. 2 cm. Merceditas, Chile. Known 1884. (292 gr., VI.) - Joe Wright, Arkansas. Found 1884. (15 gr., V; 135 gr., 3 9) ea Puquios, Chile. Found 1885. (70 gr., IV.) Mazapil, Mexico. Fell November 27, 1885. (11 gr., II.) Thunda, Queensland. Described 1886. (49 gr., VI.) Tonganoxie, Kansas. Found 1886. 40 gr. 25 cm. Algoma, Wisconsin. Found 1887. 13 gr., 8 cm. Welland, Canada. Found 1888. 85 gr. 25 cm. (11 gr., V; 9 ger., VI.) Independence, Kentucky. Found 1889. 305 gr. 42 cm. I Bridgewater, North Carolina. Described 1890. (128 gr., t IV.) Hammersley, Australia. Found 1892. (119 gr., IV.) Oroville, California. Known 1893. 18 gr. 3 cm. Plymouth, Indiana. Found 1893. (45 gr., VI.) El Capitan, New Mexico. Found 1893. 60 gr. 13 cm. 7 ee ee ee eee & po 244 BREZINA—COLLECTIONS OF METEORITES. [April 8, Tarapaca, Chile. Known 1894. (264 gr., V.) Arlington, Minnesota. Found 1894. (28 gr., IV.) Nocoleche, New South Wales. Found 1895. 84 gr. 20cm. Luis Lopez, New Mexico. Found 1896. 34 gr. 6 cm. Badger, New Mexico. Known 1897. 17 gr.5 cm. (1050 gr., IV; 5 gr., V; 971 gr., VI.) sick, Lipan Flats, Texas. Found 1897. (184 gr., V.) 57- Mollified Medium Octahedrite, Ome. Figures fallen in disorder by mollifying; points instead of Taenite lamella. 58. Coarse Octahedrites, Og. Thickness-of lamelle 1.5-2.0 mm. Tennant’s Iron. Found 1784. (7 gr., VI.) Bendego, Brazil. Found 1784. (202 gr., VI.) Bohumilitz, Bohemia. Found 1829. 54 gr. 6 cm. Wichita, Texas. Found 1836. (168 gr., II: 44 gr., V; 428 gr., V1.) | Cosby’s Creek, Tennessee. Found 1837. 22 gr. 5 cm. (37 gt., V.) Smithville, Tennessee. Found 1840. 39 gr. 7 cm. Magura, Hungary. Found 1840. (174 gr., VI.) Cranbourne, Victoria. Found 1854. (1 gr., VI.) Sarepta, Russia. Found 1854. (19 gr., IV.) Saint Francois, Missouri. Known 1863. (47 gr., VI.) Canyon City, California. Found 1872. t1oogr.,12cm. Nochtuisk, Siberia. Found 1876. 1 gr. 1 cm. Penkarring Rock, Australia. Found 1884. (2 gr., IV; 73 gt., VI.) Silver Crown, Wyoming. Found 1887. (133 gr., IV.) Azucar, Chile. Found 1887. (160 gr., IV.) Bischttibe, Russia. Found 1888. (263 gr., VI.) Cafion Diablo, New Mexico. Found 1891. (173 gr., IV; 272 gr., VI.) : Oscuro Mountain, New Mexico. Found 1895. (67 gr., IV.) Rosario, Honduras. Known 1897. (18 gr., VI.) 59. Mollified Coarse Octahedrite, Oge. Figures fallen in disorder by mollifying; points instead of Taenite lamelle. Reed City, Michigan. Found-1895. (122 gr., IV.) 60. Coarsest Octahedrite, Ogg. Thickness of lamelle 2.5 mm, and more, 1904.] BREZINA—COLLECTIONS OF METEORITES. 945 Seelasgen, Germany. Found 1847. (72 gr., VI.) Union County, Georgia. Described 1853. 11 gr. 3 cm. Nelson County, Kentucky. Found 1860. (261 gr., V; 103 gr., VI.) . Dacotah, Indian Territory. Found 1863. (90 gr., VI.) Sao Juliao, Portugal. Found 1883. 275 gr. 48 cm. (493 gr., V; 599 gr., VI.) Mount Joy, Pennsylvania. Found 1887. 57 gr. 7 cm, (29 gr.,V.) Mooranoppin, Australia. Known 1893. 26 gr. 14 cm. Arispe, Mexico. Found 1898. 19 gr. 3 cm. 61. Breccia-like Octahedrite, Netschajevo group, Obn. Me- dium Octahedrite with nuggets of Silicate. Netschajevo, Russia. Found 1846. 38 gr.10cm. (20 gr., VI.) 62. Breccia-like Octahedrite Kodaikanal group, Obk. Fine Octahedrite brecciated with nuggets of Silicate. 63. Breccia-like Octahedrite Copiapo group, Obc. Coarsest Octahedrite brecciated with Silicate-nuggets. Copiapo, Chile. Found 1863. 4 gr. 2 cm. 64. Breccia-like Octahedrite Zacatecas group, Obz. Octahe- dral nuggets breccialike with globes of Troilite. Zacatecas, Mexico. Knowns1s520. 30 gr. 10cm. Barranca Blanca, Chile. Found 1855. (17 gr., IV.) 65. Breccia-like Octahedrite Ngoureyma group, Obzg. Mol- ten and tracted Iron of the Zacatecas group. Ngoureyma, Algiers. Fell June 15, 1900. 173 gr. 40 cm. (29 gr., IV.) G. HEXAHEDRITE.—Structure and cleavage hexahedral. 66. Normal Hexahedrite. Neumann-lines ungrained. Lime Creek, Alabama. Found 1834. (8 gr., IV.) Coahuila, Mexico. Known 1837. (293 gr., VI.) Fort Duncan, Texas. Known 1852. 183 gr. 45 cm. (99 gr., VI.) Scottsville, Kentucky. Found 1867. (57 gr., VI.) DeSotoville, Alabama. Found 1878. (158 gr., IV; 851 gr., VI.) Hex River, Cape Colony. Found 1882 (216 gr., VI.) 246 BREZINA—COLLECTIONS OF METEORITES. _[April 8, Iredell, Texas. Found 1898. 6 gr. 4 cm. Murphy, North Carolina. Found 1899. 43 gr., 13 cm. 67. Grained Hexahedrite, Ha. Structure and cleavage run- ning through the whole mass consisting of Bde with differ- ently orientated sparkling. Floyd Mountain, Virginia. Found iba. (400 gr., VI.) 68. Brecciated Hexahedrite, Hb. Mass consisting of differ- ently orientated hexahedral grains. Kendall County, Texas. Known 1887. 299 gr. 41 cm. H. ATAXxITE.—Structure interrupted. 69. Capegroup, Dc. Rich in Nickel; sharp (hexahedral?) etching bands in dull mass. 70. Shingle Springs group, Dsh. Richin Nickel; not sharp parallel spots. 71. Babbsmill group, Db. Rich in Nickel; lusterless, fibene geneous mass. Deep Springs, North Carolina. Found 1846. (30 gr., VI.) 72. Linnville group, Dl. Rich in Nickel; meandering- veined or latticed. San Cristobal, Chile. Found 1882. (5 gr., V; 411 gr., VI.) Ternera, Chile. Described 1891. 1 gr. 1 cm. 73. Nedagolla group, Dn. Poor in Nickel, grained, no ridges. Rafriiti, Switzerland. Fell October, 1856. 1 gr. r cm, Forsyth County, Georgia. Found 1891. 221 gr. 37 cm. Ophir, Montana. Found 1897. 55gr.30cm. (30gr., IV.) 74. Siratic group, Ds. Poor in Nickel; ridges, incisions or enveloped Rhabdites. Campo del Cielo, Argentina. Found 1873. 8 gr. 2 cm. Chesterville, South Carolina. Found 1847. 12 gr. 6 cm. Locust Grove, Georgia. Found 1857. (206 gr., VI.) 75. Primitiva group, Dp. Poor in Nickel; silky streaks and luster. Primitiva, Chile. Found 1888, 306 gr. 44 cm. (29 gr., VI.) Plate VII, Fig. 49. 76. Muchachos group, Dm. Poor in Nickel, grained, por- phyritic with Forsterite. Telluric Iron, Tell. Ovifac, Disco, Greenland. Found 1808. (795 gr., VI.) PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLaTE I. BREZINA—METEORITES. PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLATE Il. BREZINA—METEORITES. PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLATE III. BREZINA—METEORITES. PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII. No. 176. PLATE IV. FiG. 42. FIG. 43. BREZINA—METEORITES. PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLATE V. FIG. 44. FIG. 45. BREZINA—METEORITES. PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLATE VI. FIG. 46. FIG. 47. BREZINA—METEORITEsS. Ce Ps: PROCEEDINGS AM. PHILOS. SOC., VoL. XLIII, No. 176. PLATE VII. FIG. 48. FIG. 49. BREZINA—METEORITES. 1904.] DUDLEY—PASSENGER CAR VENTILATION. » 247 Niakornak, Greenland. Found 1819. (1 gr., VI.) Sao Francisco, Brazil. Found 1874. (479 gr., V; 26 gr., VI.) Nikolojewskaja Wosimskaja, Russia. Found 1883. 87 gr. 8 cm. Artificial Products. Glowed Steel. (1 gr., VI.) Devitrified molten Pitchstone. (VI.) IX. DvupLicaTes FoR EXCHANGES. In a synoptical collection the duplicates destinated_ for exchanges should be registered separated from the pieces of the main collection by three reasons: to avoid the constant moving of weights in the catalogue, to avoid parting with specimens, which show important peculiarities and to enable directors or owners of other collections to arrange proposi- tions for exchange. | As the present article has a more theoretical scope, dupli- cates were not registered at all; they form a series of go localities in the weight of together 85 kilograms. A SYSTEM OF PASSENGER CAR VENTILATION. BY CHAS. B. DUDLEY, PH.D., CHEMIST, PENNSYLVANIA RAILROAD COMPANY. (Read April 8, 1904.) The ventilation of passenger cars is no small problem. The ordinary passenger coach includes about 4,000 cubic feet of space, and the difficulties of the problem will be apparent when it is stated that it is proposed to take into this limited space sixty people, to keep them in that space for from four to six hours at a time, to keep them warm enough for their comfort in winter, to supply them with the necessary amount of fresh air, and at the same time to, as far as possible, exclude objectionable material from without, such as smoke and cinders. It is perhaps not strange, in view of the small space and large number of people and the inclemency of the weather, that progress in the solution of the problem has been slow. It is believed that the system which will be described is a decided step forward in this matter, and while it may not be the 248 . DUDLEY—PASSENGER CAR VENTILATION. [April 8, final solution, it certainly can be justly claimed that it is a marked amelioriation of the conditions which have prevailed heretofore. Before proceeding to describe the system, it may perhaps be wise to consider briefly a few preliminary questions as follows: First, is it possible to properly ventilate a car without having the heating system a part of the ventilating system? It will be quite obvious on a moment’s reflection, we think, that in the climate in which we live unwarmed air, especially in view of the large amount of it required, as will appear later, cannot be successfully taken through so small a space as a passenger coach in sufficient quantity to properly ventilate it. Little argument is needed on this point, and in the system to be described the heating system has been regarded as an essential feature. A few proposed systems have ignored the heating system, but none of them, so far as known, can be regarded as efficient in cold weather. A second point that deserves a moment’s attention is, ‘* When can a space be said to be well ventilated, or what is the standard - of good ventilation?” It is well known that three things are con- tinually given off from the bodies of human beings which tend to make any space in which they are situated for any length of time have the characteristic which is called ‘‘ill ventilated.’’ These three things are carbonic acid, water vapor, and a certain sub- stance which for want of a better name is commonly called ‘‘ organic matter,’’ and which is believed to be the source of the odor. Of these three, carbonic acid is easily determined. Those who are familiar with the studies that have been devoted to ventilation, and which are described in standard works on Hygiene, are aware that formerly an arbitrary amount of carbonic acid in the air was taken as the measure of good ventilation. It is well known that the ordinary outside air contains about four cubic feet of carbonic acid in 10,000 of air, and formerly it was customary to say that if the carbonic acid in any closed space occupied by human beings did not exceed ten cubic feet per 10,000, the space might be regarded as well ventilated. Later studies seemed to have changed this view, | and the test that is now given in the standard works on ventilation is that a space can be said to be well ventilated when a person coming into that space from the outside air does not detect any of the odor which is characteristic of badly ventilated spaces. Quite a large number of analyses have been made to determine how much carbonic acid is characteristic of air of this kind. The best and most careful studies on this subject are probably those given 1904.) DUDLEY—PASSENGER CAR VENTILATION. 249 in Parke’s Practical Hygiene, and it is found that when the car- bonic acid naturally in the air, is increased by two cubic feet per 10,000 from human beings, it is possible to begin to detect the odor mentioned. So that when an analysis of the air in any closed space, which is occupied by human beings, shows not more than six cubic feet of carbonic acid per 10,000, it is claimed that the space may be regarded as well ventilated. The third point to be discussed is, since carbonic acid is given off from human beings, and since the amount of it in the air from this source is an essential element in ventilation, it is necessary to know how much carbonic acid per person per hour goes into the air. The same authority already quoted, namely, Parke’s Practical Hygiene, gives the results of a very large number of experiments on this subject. Men usually give off more than women, and children less than either. A man in vigorous work gives off more than in idleness. The mean of a mixed community, such as may be assumed to ride on cars, is 0.60 cubic foot per person per hour. It will be seen in a moment where these figures apply. Fourth, one of the most important questions in car ventilation is, ‘‘ How much air per car per hour is it necessary to take through a car in order to have it well ventilated?’’ If the data above given are to be trusted, it will be evident that when a car contains sixty people, each one giving off on the average 0.60 cubic foot of carbonic acid per hour, there will be per hour thirty-six cubic feet of carbonic acid to deal with, and the problem becomes how much fresh air is it essential to mix with these thirty-six cubic feet of carbonic acid, in order that it may be diluted to such an extent that it will not add to the carbonic acid already in the air, more than two cubic feet per 10,000. The problem may be stated in the form of proportion. If 10,000 cubic feet of air are to contain two cubic feet of carbonic acid in addition. to its normal amount how many cubic feet are essential to contain thirty-six? Reducing the proportion and the astounding figure is reached that, in order to have a passenger car well ventilated according to the data already given, it is essential to take through the car 180,000 cubic feet of fresh air per hour. This figure may be stated in another way, and this is the form in which it is usually given in treatises on ventila- tion, namely, it requires 3,000 cubic feet of fresh air per person per hour to maintain the air in any closed space in the conditions ‘required for good ventilation, according to the standards already mentioned. ‘This is hardly the place to discuss the validity of this 250 DUDLEY—PASSENGER CAR VENTILATION. [April 8, figure. There is some difference of opinion as to whether the amount is not excessive, and it is perhaps fair to say that the point cannot be regarded as satisfactorily settled. It may not be amiss to mention that in conversation with Professor Atwater, of the Wesleyan University, Middletown, Conn., who has made a large number of experiments in the human calorimeter, he stated that the inmates of the calorimeter do not complain of drowsiness or of any unpleasant feeling even though the carbonic acid reaches a very much higher figure than anything that has been mentioned, but do complain of drowsiness and languor with occasional headache if the amount of moisture in the air gets much above the normal. It may not be amiss to mention at this point that, in the system of car ventilation to-be described, no attempt has been made to get the large amount of air mentioned, namely, 180,000 cubic feet per car per hour. The experimental work has been limited to an attempt to get 60,000 cubic feet of air per car per hour, or about 1,000 cubic feet of fresh air per person per hour. Fifth. One more question must be discussed a little in order that what follows may be completely understood, namely, how is it possible to measure the amount of air that goes into and out of a car per hour? It may be said that attempts to do this by means of sizes of apertures and velocity of currents have not succeeded very well, and it will be obvious why this should be so, since there are a very large number of small apertures in a car, both inlets and out- lets, all of which are elements in the problem. No window or door is tight, and even though the velocity of the air going out of the ventilators is measured, the friction against the sides of the venti- lators is such that it is very difficult to get an average figure for ve- locity. Accordingly another method has been employed, as follows : A car is loaded with a definite number of inmates, and after a run under ordinary conditions a sample of the air from the car is secured and analyzed for carbonic acid. It may be supposed that the analysis shows twelve cubic feet of carbonic acid per 10,000 cubic feet of air. But four cubic feet are a normal constituent of the air, leaving eight cubic feet coming from the inmates of the car. If there were sixty of these and each one gives off carbonic acid at the rate of 0.60 of a cubic foot per hour, it is obvious thirty-six cubic feet per hour are to be dealt with and the problem becomes, ‘‘ How much fresh air must be mixed with thirty-six cubic feet of carbonic acid from the car inmates, so that the resulting mixture would show on analysis eight cubic feet of carbonic acid from the same source 1904.] DUDLEY—PASSENGER CAR VENTILATION. 251 per 10,000 of the mixture ?’’ or in other words, if 10,000 cubic feet of air contains eight cubic feet of carbonic acid, how many cubic feet of air will be required to contain thirty-six cubic feet of car- bonic acid on the same ratio? Reducing the proportion and it appears that under the conditions assumed, 45,000 cubic feet of fresh air per hour would pass through the car. In this illustration the amount of air in the car to start with is ignored, since it is not 2 i in 00 | ool == Se — =_, [o) t - + _ . i \ T \ \ ) 1 : oo pa 1 OO} va an important element, and since results accurate to a few hundred cubic feet are not essential. This brings us to the consideration of the ventilating system which has been adopted as standard by the Pennsylvania Railroad Company, and which is now in daily use on some 800 passenger coaches on the lines east of Pittsburg and Erie, and on nearly 200 coaches on the lines west of Pittsburg and Erie. It may not be amiss to say that the development of the system has taken several years of study. Avery large number of experiments have been 252 DUDLEY—PASSENGER CAR VENTILATION. [April 8, made. Each experiment led to modifications and changes, which were followed often by runs on the road, with analyses of air from the cars, the information obtained from each trial being used to lead to further modification, until a satisfactory result was obtained. Of course, the system as described is the one finally decided on. It will be noted by an examination of the plate that, taken as a whole, the system is extremely simple. It consists in taking air from the outside in through hoods located on what is known as the lower deck near the top of the car, and at diagonally opposite corners of the car. The opening of the hoods is toward the direction in which the car is moving, and, as will be noted a little later, the movement of the car is an element in the problem of getting the air into the car. The opening is covered with a gauze to exclude large cinders. The hood is fitted with a valve operated from the inside of the car in such a way that as the car changes direction the proper passageway is provided. From the hood the air passes through what is technically called a ‘¢ down-take,’’ about 100 square inches in area, which conducts the air down underneath the floor of the car to a passageway which occupies the space between the out- side sill, the floor, the first intermediate sill and the false bottom. This space has an area of a little over 100 square inches, and extends the whole length of the car. From this space the air passes up through apertures in the floor into the heater boxes, where it is warmed. From the heater boxes the air passes out into tubes situated under each seat, and is delivered into the aisle of the car from the tubes. From these points it disseminates through the car, and finally passes out of the car through ventilators situated along the center line of the upper deck. These ventilators may be of any approved form. The kind most used thus far has been what is known as the ‘*Globe Ventilator,’’ which ventilator has the characteristic that when the car is moving through the air, or when the wind is blowing across the ventilator, a suction is produced on the air in the car, It will be observed from what has already been stated, that there are two things that cause the ventilating air to move through the car. First, the heating system. The ventilators in the upper deck are situated some two feet higher than the top of the hood, and accordingly, when there is heat in the car, or when the lamps are lighted, there is the proper ventilating movement of air through the car, due to this force. Also it will be noted that the movement of the car is an element in the problem likewise. The car movement produces pressure in the hood and down-take and the 1904. DUDLEY—PASSENGER CAR VENTILATION, 253 ventilators produce suction as has already been described, and these two acting together lead to change of air in the car. The control of the system, that is the means by which the amount of air passing through the car is diminished, is in the ventilators. Each venti- lator is provided with a register, and when these registers are closed the total amount of air passing through the car is diminished a little over one-half. It will be observed that by partly closing all the ventilators, or closing a part of them, any intermediate figure between these two can be obtained. It is found essential to have a portion of the passageway in the ventilators over the lamps contin- uously open, in order to carry off the products of combustion. The register at no time closes this portion of the ventilators over the lamps. The experimental work having led up to the construction finally decided on, it remained to actually put the system to test and see exactly what was being obtained. The first experiment was to see whether when the car was standing still, and heat was in the car, the movement of the air would be in the desired direction, namely, into the car through the hood, and out through the ventilators on the upper deck. With some systems of car ventilation where the movement of the air is almost wholly a function of the movement of the train, when the train stops the air movement is in the opposite direction, owing to the relation between the heating system and the ventilating system. In the system which we are dealing with this does not take place, for obvious reasons, namely, as has already been stated, the exits are higher than the in-takes. The second test concerned the air from the closet. Some anxiety was felt as to whether the ventilating system would take air from the closet into the car. As a precautionary measure a small ‘*Globe’’ ventidator was put in the roof of the closet, and also the proportions of parts of the system were designed in such a way that when the car was in service, there would be a plenum in the car produced by the hood rather than a vacuum produced by the suction of the ‘*Globe’’ ventilators. As a matter of fact, the construction finally adopted gave very close to a balance between these two features. However, many experiments show without question that when the car is in motion, and the ventilating system in full operation, the air movement is toward the closet, instead of from it. A third test of the system was to determine the actual amount of air passing through the car. In order to decide this question, a car 254 DUDLEY—PASSENGER CAR VENTILATION. [April 8, fitted as above described was filled with men from the shops, who were paid for their time, under the charge of a foreman, so that they could be controlled in the matter of opening doors and windows, and a trip was made early in December from Altoona to Johnstown and return. Rubber bags and hand bellows were taken along with which to secure samples of the airin the car. Steam heat was necessary since the temperature outside was from 23 to 30 degrees Fahrenheit, and neither door nor window was opened dur- ing the trip, except that after the proper samples of air had been taken at Johnstown the men were allowed some freedom, since a wait of a couple of hours must ensue before the return trip could be made. The air samples for analysis were taken by pumping air into the rubber bags by means of the hand bellows, moving from one end of the car and back again in the aisle during the operation, and taking the air from about the level of the heads of the passengers. The analyses were made immediately after the return. In making the air analyses the carbonic acid only was determined, and from this was calculated the amount of fresh air taken through the car per hour by the ventilating system, the method used being the one described earlier in this article. As a matter of fact, there were fifty-two men in the car, and being workingmen it was assumed that they gave off 0.72 of a cubic foot of carbonic acid each per hour. The figures obtained on the trip mentioned are as follows: WESTBOUND. Per cent, of | Cubic feet of carbonic air per car acid per hour All Globe ventilators closed—Bennington ..... 0.18 ° 26,700 All Globe ventilators open—Buttermilk Falls . , 0.10 62,400 All Globe ventilators open—standing twenty minutes at Johnstown ,........... aeecess 0,21 22,000 EASTBOUND. All Globe ventilators closed—Cresson,........ O.14 7,400 All Globe ventilators open—McGarvey,...... 0,10 2,400 All Globe ventilators open—standing twenty minutes at Altoona .......... jee kee he's 0,20 23,400 1904.) DUDLEY—PASSENGER CAR VENTILATION, 255 In explanation of the figures it may be stated that the stations mentioned denote locations at which air samples were taken. Bennington, on the schedule used, is about twenty-three minutes from Altoona; Buttermilk Falls is about fifty-seven minutes from Bennington, and Johnstown is about ten minutes from Buttermilk Falls. Returning, Cresson is about forty-two minutes from Johns- town; McGarvey about twenty minutes from Cresson, and Altoona about five minutes from McGarvey. These figures will give some idea of the interval between samples. As has already been stated, the system was designed to give 60,000 cubic feet of air per car per hour, and it was felt that the figures given above show that the system fairly well fulfills the pur- pose for which it was designed. Not more than 60,000 cubic feet were planned for, for the reason that it was found impossible, as the result of experiments made early in the studies on this subject, to warm the large amount of air required by theory. While it would perhaps be possible to warm more than 60,000 cubic feet of air, yet it is always desirable to have some little factor of safety in the appliances used, and accordingly, after very careful consultation over the matter, it was decided not to attempt to get more than 60,000 cubic feet per car per hour. Two points further were made the subject of test: First, the ability to keep the cars warm, even in the severest weather. This with the heating system, for which the ventilating system was designed, was found to be extremely satisfactory. Careful obser- vations were made both on long runs and during severe cold blizzards on this point by competent persons, and at no time has there been any difficulty in keeping the car comfortable. Further- more, the distribution of the heat in the car seems to be entirely satisfactory. Even under the influence of severe winds, not more than two or three degrees difference in different parts of the car are observable. It may be worth while to mention that, as will be noted from the description, the ventilating system consists of two halves, which are entirely independent of each other, except that the heating system on the two sides takes steam from the same point. Careful experiments with cars on the road indicate that when the wind is directly ahead, the two sides of the car take in approximately equal amounts of fresh air. When the wind, how- ever, is to the right of the line of motion of the car, that side of the car seems to do most of the ventilating, and when it is to the 256 DUDLEY —PASSENGER CAR VENTILATION. (April 8, left, that side does the most of the ventilating. This will be readily understood from the construction. The curvature of rail- roads, however, is so great that this fluctuation in the amount of fresh air taken in on the two sides does not, as already stated, seriously affect the temperature in different parts of the car. The final test made has been as to the ability of the ventilating system to exclude objectionable matter, such as smoke, cinders and dust. A good deal of interest was felt over this matter at the start, and it is to be confessed that anything which is fine enough to be carried in the air will ultimately find its way into the car, As a matter of fact, it is found that small cinders which pass the gauze on the hood of the in-take are distributed more or less along the bottom of the space underneath the floor, and it occasionally becomes essential to clean these cinders out. Also in going through tunnels, sometimes smoke and gases are taken into the car in small amount. To meet this difficulty a butterfly valve was put in the down-take, and the instructions provide that this shall be closed when going through tunnels. Furthermore, the air being taken from near the top of the car, dust rarely gets high enough to cause any trouble. Smoke from the locomotive usually is either diverted by the wind, or is high enough not to reach the in-takes, so that, as a matter of fact, less difficulty has been experienced from objection- able material going in along with the fresh air than was feared. Finally, the air in the car being completely changed once in four minutes, it is evident that the inconvenience from temporary foul air going into the in-takes is reduced to a minimum. The system as described has been in daily use on more or less cars, for now some five years, and the criticisms leading to modifi- cations have been less than might have been expected. The system is being applied to all new cars as fast as they are built, and to other cars in the equipment as fast as circumstances will admit. It is unfortunate to be compelled to say that the system has not yet been applied to a sleeping car. Altoona, Penna., April 7, 1904. PROCEEDINGS AMERICAN PHILOSOPHICAL SOCIETY HELD AT PHILADELPHIA FOR PROMOTING USEFUL KNOWLEDGE Vou. XLIII. Aprit-SEPTEMBER, 1904. No. 177. THE ORIGIN AND NATURE OF COLOR IN PLANTS. BY HENRY KRAEMER. (Read April 8, 1904.) A list of the more important papers published, up until within the past ten years, on the subject of plant colors is given in Dippel’s Das Mikroskop.' Of these the papers by Pringsheim’ on the examination of chlorophyl and related substances, and by Miiller* on the spectrum-analysis of the color substances of flowers, are probably the most important. Pringsheim confined his attention mainly to a spectroscopic study of chlorophyl and the yellow substances in germinating plants, yellow flowers and yellow autumn leaves. He concluded that the yellow substances from these several sources were but modifications of chlorophyl. The yellow principle found in ger- minating plants he regarded as closely related to chlorophyl, and the yellow substance in autumn leaves as a more remote modifica- tion of it. He did not consider, however, as subsequent writers have claimed, that these substances were identical. Two years before the appearance of Pringsheim’s paper, Kraus‘ stated that he had separated from an alcoholic solution of chloro- phyl by means of benzol two distinct substances, one yellow and the other blue, the latter being taken up by the benzol. Pring- sheim, however, showed that the blue substance was in reality chlorophyl, and that the alcoholic solution, which showed faint chlorophyl-like bands in the spectroscope, still contained some chlorophyl. PROC. AMER. PHILOS. SOC. XLIII. 177. Q. PRINTED JULY 27, 1904 258 KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, While Pringsheim believed that there were two modifications of chlorophyl, one yellow and the other green, the former predom- inating in germinating plants grown in the dark, and the latter or green substance in leaves exposed to the light, still he did not believe that they could be separated from each other by the method proposed by Kraus. Yet notwithstanding Pringsheim’s well-founded criticisms of the method employed by Kraus, and taking for granted that there were two principles composing chlorophyl, nearly all investigators since Kraus’s work was published have practically employed his method as modified by Hansen ® for the separation of the so-called yellow and green chlorophyl. According to this method of Hansen, fresh material is extracted with 95 per cent. alcohol, the liquid filtered, and to the filtrate 30 to 50 per cent. of water is added ; the solution is shaken with petroleum ether and the. liquids sepa- rated, the ether taking up the green substance, or chlorophyl proper, and the hydro-alcoholic solution holding the yellow prin- ciple. If autumn leaves are treated in the same way, the ether solution will contain very little chlorophyl, while the hydro-alcoholic solu- tion will contain a yellowish or reddish substance, depending upon the kind of material examined. It has usually been con- sidered that this yellow substance in autumn leaves is associated in summer with the active plastids, and on account of its having little food value remains behind. It has furthermore been considered by many that the yellow principle in young leaves is identical with that in autumn leaves and the yellow substance found in yellow flowers, fruits and roots. Kinps or COLORS IN PLANTs. Colors in plants may be considered to be due to definite con- stituents which either themselves are colored or produce colors when acted upon by other substances. ‘These substances are found in all parts of the plant, and apparently in all of the cells excepting certain meristematic or dividing cells. ‘They may be divided into two well-differentiated classes, namely, (1) those which are associ- ated with the plastids, or organized bodies in the cell, and (2). those which occur in the cell-sap, or liquid of the cell, Le a Pe I eee ae ee Ee 1904.] KRAEMER—NATURE OF COLOR IN PLANTS. 259 SO-CALLED WuitE Cotors. The so-called white colors in plants do not properly belong to either class, but may be said to be appearances rather, due to the absence of color, and depending upon the reflection of light from transparent cells separated by relatively large intercellular spaces containing air. In other words the effect produced by these cells may be likened to that produced by the globules in an emulsion. The white appearance is most pronounced in the pith cells of roots and stems, where on the death of the cells the size of the inter- cellular spaces is increased and the colorless bodies in the cells as well as the walls reflect the light like snow crystals. METHODS OF EXTRACTION. During this investigation I have examined by means of the Leitz micro-spectroscope the various kinds of coloring substances to which I shall refer but, except in the case of chlorophyl, did not obtain results which were entirely satisfactory, and will endeavor to give special attention to this phase of the subject in another paper. It is frequently difficult to extract and isolate these substances in a sufficiently pure condition for spectroscopic work, particularly as many of them change rapidly. In this paper, therefore, I shall confine myself to the considera- tion of the behavior of the extracted coloring substances toward chemical reagents. The material containing the coloring matter was in all cases. separated as nearly as possible from that which was free from color or contained it in less amount. Various solvents were used in the extraction of the coloring substances, depending upon the solubility or nature of the substance. The solvent mostly employed was alcohol (95 per cent.), in some cases dilute alcohol (50 per cent.) or water (hot or cold) was employed. The plastid colors were extracted by placing the fresh material in 95 per cent. alcohol and allowing it to macerate in the dark for a day or two. I usually took the precaution to tear the material with the fingers rather than to cut it. The solution so obtained contains other than the plastid coloring substances, which latter may be isolated in a more or less pure condition by either of the following methods: (1) The alcohol is distilled off and the solution evaporated on a water bath to near dryness, boiling water is then added and 260 KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, the solution filtered, the extract washed with hot water until the filtrate is colorless ; the-extract is then taken up with cold alcohol. (2) In the other method the alcoholic solution is diluted with water ; and ether, benzin, benzol, xylol, or other similar solvent is added, and the mixture shaken in a separatory funnel. The ethereal layer containing the plastid color may be further purified by shak- ing it in a separatory funnel with alcohol, adding sufficient water to cause separation of the two layers. The ethereal solution is then distilled and evaporated on a water bath to near dryness, and the pigment taken up with cold alcohol. In either case the alcoholic solution may be boiled for an hour or two with zinc in a reflux con- denser, whereby the more or less oxidized plastid pigments are restored. This is a particularly important procedure in the micro- spectroscopic examination of chlorophyl, and may be used as a means of detecting chlorophyl in other substances. In order to obtain the coloring principles in early leaves, as the red coloring principle in the leaves of oak, rose, etc., it was found most satisfactory to extract the material with alcohol, add xylol or similar solvent, and then sufficient water to effect separation of the solutions, using a separatory funnel. The cell-sap color remains in the hydro-alcoholic solution, and the traces of xylol should be removed by heating the solution on a water-bath, as the presence of xylol causes a cloudiness in the solution on the addition of the re- agents to be subsequently employed. The cell-sap colors of flowers, as of pansy, tulip, etc., are sepa- rated from the plastid pigments in the same way as just mentioned in connection with early leaves. ‘ The cell-sap colors in fall leaves are easily removed by treating the more or less comminuted material with hot or cold water. In some cases there are several associated colors, and these may be extracted separately by taking advantage of their varying solu- bility, as in the case of carthamus, where the red principle is ex- tracted with water and the yellow principle with alcohol. In still other cases special methods are employed, as in the ex- traction of carotin from carrot according to the method proposed by Husemann.* The grated carrot is mixed with water, squeezed through cheese-cloth, and a small quantity of dilute sulphuric acid and tannin added to the mixture, forming a coagulum which settles to the bottom of the precipitating jar. The supernatant liquid is removed by means of a syphon and the coagulum treated six or 7 1904.] KRAEMER—NATURE OF COLOR IN PLANTS. 261 seven times with 80 per cent. alcohol, which removes mannit and hydro-carotin; the coagulum is then extracted with hot carbon disulphide, which removes the carotin. This solution is evaporated to about half the original volume, an equal amount of absolute alcohol added, and set aside to crystallize, the carotin separating. One of the striking observations made during this investigation was that in the case of the cell-sap colors the solution was different in color, as compared to the natural color, or sometimes almost colorless, reagents, however, striking colors as intense or even more intense than the original colors. For the convenience of those who may wish to follow similar studies, the plants which I examined may be grouped according to the solvents which I found best adapted for the extraction of the coloring substances. There is also given the part of the plant employed and the color of the solutions I obtained. CoLoR PRINCIPLES EXTRACTED WITH ALCOHOL. Name oj Plant. Part Used. Color of Solution. 1. Apple (Baldwin) (P Malus) . carp ht yellowish-red 2. Apple (Bellefieur) (Piyrus Malus). . |Epicarp ‘Pale yellow 8. Arbutus (Epigea repens) . . . . |Petals fee straw 4. Azalea (Azalea nudiflora) - Bike ts ate hits Petals ‘Pale straw . 5. Beet (Beta vulgaris)... 2... 2 se Leaves \Deep green 6. Black Rubus Canadensis) . . ‘Stems Reddish-brown 7. Ba meronp (f anunculus acris). .... . Petals ‘Deep yellow 8. Cabbage, red (Brassica oleracea) Leaves |Purplish-red 9, Capsicum (Copetoum my fata) : |Dried fruit Yellowish-red 10. Carnation, red aryophyllus) Petals Deep red 11. Carrot Sse Carota) ene Root Deep reddish-yellow 12. Celery (Apium graveolens)....... Etiolated leaves ae t gre h-yel- ; ow : 13. Chondrus (Chondrus crispus)...... Fronds Light yellowish- green 14. Cinquefoil (Potentilla Canadensis) . . . |Petals |Greenish-yellow 15. Cran (Oxycoccus . |Fruit |Deep red 16. Daffodil (Narcissus Pseudo-Nareissus) « Petals Deep yellow 17. Dandelion (Taraxacum officinale) . Petals Lemon-yellow Dock (Rumez crispus)........+.. - Spring leaves Reddish-brown 19, is (Cornus Florida)..... an it Brownish-yellow 20. Dulce (Rhodymenia palmata) ...... Fronds Light yellowish- green 21. Elder (Sambucus Canadensis)...... ~ oy leaves Reddish-brown 22. Fucus (Fucus vesiculosus). ....... mds Greenish-brown 23. Hepatica (Hepatica triloba)..... . . |Petals Lemon-yellow or greenish-yellow 28a. Hepatica (Hepaticatriloba)....... Involucre Jish- 24, (Irie veravcolor) .. 1. .52 22. Peta: Violet 25. Jack-in-the-pul pit (Arisema triphyllum) Spathe lish 26. Japanese quince (Cydonia Japonica) . etals t purplish-red Wha. LEMON DEO)... ja eae ack i's . . . . |Epicarp be 27. Mallow Malva UMOPLE aso ns moist Petals Viol 28. Maple (Acer rubrum) ......¢...-.. Flowers Yellowish or brown- 29. merems (Calendula officinalis)... . . |Petals at yellow 30. Oak, red (Quercus coccinea?).....- Spring leaves Red -brown 30a. Orange peel pikes, a0 jepIcarp eenge rege 31. Pansy, bine (Viola tricolor) She aay Say 2 Petals 'Purplish-red 262 KRAEMER—NA'TURE OF COLOR IN PLANTS. [April 8. CoLoR PRINCIPLES EXTRACTED WITH ALCOHOL— Continued. Name oj Plant. Part Used.. Color of Solution. 32. Pansy, arsine (Viola tricolor,,..... Petals ues llow 33. Pinea ar e@ (Ananas sativa) ....... Outer portion te De 34. (Raphanus Raphanistrum) . Purplish layer of root Tight red 35. Rose (Resa gallica) ......-22.-. : |Dried a Light brown 35a. Rose (Rosa er? ata ur rayne = Early leaves Reddish-brown 36. Safflower (Carthamus tinctorius).... . Petals Rees Ye yellow 37. Saffron ( sativus) . |Dried stigmas llowish-red 38. Skunk cabbage Bpatkyeme arr . . Green leaves Deep green 39. Skunk cabbage (Spathyema fetida . Inner portion of leaf|Deep yellow - ‘ 40. Skunk cabbage (Spathyema fetida) . |\Spathe (buds! Deep yellowish-red 41. Skunk cabbage Yemone joan} . . \Scales Purplish-red 42. Skunk cabbage (Spathyema fetida . |Tips of leaf buds t Src nti 43. Spinach bed cowry rats EOF Hi Leaves Deep greén. 44. Sweet Cicely (Washingtonia Cmeneneyst Spring leaves Reddish brow 45. Tomato ( sicon esculentum)... . |Fruit Pale yellow 46. Tulip pyc oe Apaalane : Petals ht reddish-brown 47. Turnip (Brassica napus). ........ Purplish layer of root) Pale yellow 48. Violet, blae (Viola ¢ Ae Tea i ies Petals es 49. Violet, yellow (Viola scabriuscula) . . . |Petals Yellow 50. Wahoo (Euonymus Americanus) . . |Winter leaves Reddish-brown CoLoR PRINCIPLES EXTRACTED WITH DILUTE ALCOHOL. . Black Mexican corn (Zea Mays) . Geranium, house (Pelargonium -) Geranium, wild (Geranium maculatum) Hyacinth, dicustonte cerulea (Muscari yoides) nth, blue (Muscari botryoides) . . vulgaris H Strawberry (Fragaria — 4 er blue ere euculata) a Tahee eae Wistaria (Kraunhia frutescens) . . (Grains Light purplish-red Petals Light purplish-red Petals ag 7 preg Petals Petals Light Senowiih. -red Petals Purplish-red Petals Brownish-yellow gee portion of pe-|Pale red Fruit Yellowish-red Petals Greenish-yellow Petals Pale brown CoLor PRINCIPLES EXTRACTED WITH WATER. , i ‘a 69. a nis PG) «ve ewe 70. mony Tlex A lolium) . pics asl rangea ; Triien incl Lon Virginiana). 73. Mallow (Malva . Maple (Acer aace » Marigo id (Calendula oft Kishin + ¥ white (Quercus £5! 5 drhmcsate Rhubarb b (Rheum . Rose, wild (Rosa ———) + ee ar . Bafllower (Carthamua tinctoriua), . Saffron (Crocus sativus) . Solomon's seal (Vagnera racemosa) . Ye tb Gy) ie, et ee Autumn leaves ers agen oot Outer portion of Bro pniaki red stems | Fruit ei ae Fruit p red Autumn leaves Redaish-brown y n Purplish . . |\Dried fruit Purplish-red . |Frait Purplish-red Fruit Deep brownish-red . |Neutral flowers Brownish-red Autumn leaves Deep brownish-red Petal Dark purplish-red Autumn leaves Brownish-red tals Deep brownish-red . Autumn leaves Brownish-red Outer portion of pe- ae Pale red Deep brownish-red : Dried Petals Deep brownish-red . |Dried pigmas peep yellowish-red , | Fruit Deep red 1904.] KRAEMER—NATURE OF COLOR IN PLANTS. 263 PLastip COLOR SUBSTANCES. The green color in plants is due, as is well known by botanists, to a green pigment known as chlorophyl which is associated with a plastid or organized protoplasmic body, forming a so-called chloroplast. Chlorophyl is distinguished from all other plant sub- stances by possessing a dark broad band between the Fraunhofer lines A and C at the red end of the spectrum, which is apparent even in very dilute solutions. It also shows in more concentrated solutions a broad band extending from F to the violet end of the spectrum, a narrow band between C and D, or the orange portion of the spectrum, and two narrow bands between D and E, or the yellow portion of the spectrum. Pringsheim examined spectroscopically solutions of the yellow substances found in etiolated germinating leaves, and also the yellow substances of yellow flowers and autumn leaves, and observed the characteristic chlorophyl bands only by using tubes more than three hundred millimeters thick. Inasmuch as small tubes holding five or ten cubic centimeters are sufficient for the examination of chlorophyl, by means of the Zeiss or Leitz micro- spectroscope, and also because a dilute solution is necessary, one is surprised that Pringsheim and others have used tubes of such enormous thickness, and that they concluded from the more or less indistinct bands which they observed that these substances were modifications of chlorophyl. It is not at all unlikely that what he actually had were concentrated solutions of as many different principles, each of which contained traces of chlorophyl, notwith- standing the care he exercised in separating the green and yellow portions in the material which he used. In my own studies on the yellow principle of developing leaves I used the buds of skunk cabbage, which develop under ground and under leaves and are of considerable size before exposed to light. The outer light greenish-yellow portions were removed, and only the intense yellow central portion used. This material was extracted in the dark with alcohol. The solution thus obtained is of a pure lemon-yellow color, and may be freed from cell-sap sub- stances either by evaporation to an extract, washing with water, dissolving in cold alcohol, and then boiling with zinc ; or by treat- ing the original alcoholic solution with petroleum benzin, whereby the pure yellow leaf substance is separated from the cell-sap substance. 264 KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, This yellow principle is combined with plastids, which are about one micron in diameter, being spherical or polygonal in shape, and lying closely packed in the palisade cells of both the upper and lower surfaces of the leaf. The yellow plastids are distinguished from the leucoplastids, which occur in the epidermal and mesophyl cells, as well as the chloroplastids, which are found later in the green leaves, by being smaller, relatively more numerous and by not manufacturing either reserve or assimilation starch. The associated pigment is further distinguished from chlorophyl by not being fluorescent; in having a broad band extending from 65 to the red end of the spectrum, and another extending from 50-52 to the violet end of the spectrum, when examined by means of the Leitz micro-spectroscope ; and in being less soluble in alcohol and more so in benzin than chlorophyl. This latter characteristic affords a means of partially separating it from chlorophyl, and for this principle I propose the name efiopAy/, and for the associated plastid, which seems to be a distii.ct body, I propose a correspond- ing name, ef/op/ast, these terms being used expressly for the purpose of avoiding confusion. The etioplasts completely pack the cells in which they are found, and may be regarded as meristematic plastids, which later give rise to the chloroplastids. The yellow color in certain roots, flowers and fruits is apparently in all cases due to a yellow pigment associated with a plastid known as achromoplast. These plastids are distinguished from the other plastids by being of variable shape and in usually containing protein grains. The associated pigment resembles in some respects etiophyl and chlorophyl, in that it is more or less soluble in ether, benzol, xylol, carbon disulphide, etc. These pigments, for the most part, appear to be unaffected by either mineral or organic acids, but usually give some shade of green with alkalies, potassium cyanide, sodium phosphate or iron salts. In some cases they are affected by alum, iodine, sodium nitrite, or sodium nitrite and sulphuric acid, as given in Table I.’ ‘In the examination of plant colors the following reagents were found useful: Sulphuric acid, to per cent.; hydrochloric acid, Io per cent.; nitric acid, 10 per cent.; citric acid, § per cent.; oxalic acid, 5 per cent.; sodium hydrate, 10 per cent.; ammonium hydrate, 10 per cent; potassium cyanide, 1 per cent.; sodium phosphate, 5 per cent.; ferric chloride, 3 per cent.; ferrous sulphate, 2.5 per cent.; hydrogen peroxide, 3 per cent.; salicylic acid, saturated solution, gallic acid, 1 per cent.; sodium nitrite, 1 per cent; sodium nitrite followed by sulphuric 1904,] KRAEMER—NATURE OF COLOR IN PLANTS. 265 Inasmuch as there seems to be a class of these principles which are distinguished by their solubility, as well as reactions with various chemicals, I venture to propose the name chromophy/ for these yellowish or orange-colored pigments. All of the coloring substances given in Table I areYsoluble in xylol, ether and similar solvents, as well as alcohol, but are spar- ingly soluble in water. There are several substances which behave much like the plastid substances, but which are insoluble in xylol, ether, etc., and appear to occupy an intermediate position between the true plastid color substances and the cell-sap colors. I have therefore placed them in class by themselves in Table II. CELL-SAP COLOR SUBSTANCES. During the course of metabolism the plant cell manufactures other color substances which are not combined with the protoplasm, but which are contained in the cell-sap, or liquid of the cell. These substances, unlike the plastid colors, are insoluble in xylol, ether and similar solvents, but are soluble in water and alcohol, which affords a means of separating them from the plastid colors. These cell-sap pigments may occur in cells free from plastids or in the vacuoles of cells containing plastids, but not associated with them as a part of the organized body or plastid. They are usually extracted along with the chlorophyl and remain in the hydro- alcoholic solution after separation of the plastid pigment by means of xylol or other solvent. These pigments have one property in common with the chromophyl substances, namely, with alkalies, potassium cyanide and sodium phosphate, they assume some shade of green. They are distinguished, however, by the fact that the colors are markedly affected by acids and alkalies and by iron salts. They are in most cases also affected by other reagents, as shown in the accompanying tables. These substances being so sensitive to reagents, probably accounts for the various shades and tints characteristic not only of flowers but of leaves as well. My obser- vations on the germinating kernels of black Mexican corn show that even in contiguous cells the constituents associated with the dye acid; potash alum, 10 per cent.; ammonio-ferric alum, 5 per cent.; iodine solution containing .1 per cent. iodine and 0.5 per cent. potassium iodide; tannin, 3 per cent. 266 KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, vary to such an extent that the pigment in one cell is colored red- dish, in another bluish-green, and in another purplish. The results of the examination of the cell-sap colors are given in Tables III, IV and V, and while it might seem a very easy matter to divide plant colors into reds, blues and purples, it will be seen that this is almost impracticable, and that the colors given in these tables merge into one another. An examination of the color substances found in early spring leaves and in autumn leaves showed that these substances are in the nature of cell-sap colors, behaving toward reagents much like the cell-sap colors of flowers, and indeed in some instances they are apparently identical, as will be seen by comparing the results given in Table VI with those given in Tables III, IV and V. CONCLUSIONS. 1. The white appearance in flowers and other parts of plants is due to the reflection and refraction of light in more or less color- less cells separated usually by large intercellular spaces containing air. 2. The green color of plants is due to a distinct pigment, chloro- phyl, contained in a chloroplastid, and appears to be more or less constant in composition in all plants. The chloroplastid is furthermore characterized by usually containing starch. 3. The yellow color substance in roots, flowers and fruits is due to a pigment, to which I have given the name chromophyl. This substance is contained in a chromoplastid which varies consider- ably in shape, and usually contains proteid substances in addition. 4. In the inner protected leaf-buds there is a yellow principle which I have termed etiophyl, and which is contained in an organized body which I have termed an etioplast. The etioplast does not appear to contain either starch or proteid substances. 5. The blue, purple and red color substances in flowers are dissolved in the cell-sap, and are distinguished for the most part from the plastid colors by being insoluble in ether, xylol, benzol, chloroform, carbon disulphide and similar solvents, but soluble in water or alcohol. While quite sensitive to reagents yet none of these colors behave precisely alike. 6. Cell-sap color substances corresponding to the cell-sap colors of flowers are also found in early or spring leaves and in autumn leaves. 1904, J KRAEMER—NATURE OF COLOR IN PLANTS. 267 In addition I desire to say that I am inclined to look upon the chromoplastids of both flowers and fruits as having the special function of manufacturing or storing nitrogenous food materials, for the use of the developing embryo or developing seed, particularly as protein grains are usually contained in them. The same may be said of the chromoplasts in roots, as in carrot, where the pro- teids of the chromoplasts are utilized by the plant of the second year. I am further inclined to consider the cell-sap colors, like other unorganized cell-contents, as alkaloids, volatile oils, etc., to be incident to physiological activity, and of secondary importance in the attraction of insects for the fertilization of the flower and dis- persal of the seed. Finally, I acknowledge my indebtedness to Miss Florence Yaple, Philadelphia, for valuable assistance in the preparation of this paper. BIBLIOGRAPHY. 1, DippEL: Das Mikroskop, 2te Auflage, zweiter Theil, erste Abtheilung,- pages 65 and 66, 2, PRINGSHEIM: “ Untersuchungen iiber das Chlorophyll,” Monatsberichte der Koniglich Preussischen Akademie der Wissenschaften zu Berlin, 1874, p. 628. 3. MUELLER: “Spectralanalyse der Bliithenfarben,” Pringsheim’s Jahrbuch, Bd, XX (1889). 4. Kraus: “Ueber die Bestandtheile der Chlorophyllfarbstoffe und ihrer Ver- wandten,” Sitaungsdber. d. med, phys. Gesellsch. in Halle (1871). 5. HANSEN: “Fardstoffe des Chlorophylils,” 1888, quoted by Dippel. 6. HUSEMANN, A.: Die Phlansenstoffe, 2te Auflage, p. 959. KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, 268 cl " *MOTINTOS JO OOo [BUTS “9 “Ot : ‘ *JOYOoTS poyn[Ip Youu Ur s[qnyjosut st ajdjoupid Zupi0j09 ay} JY} 39By 9y} 0} JO a[droutid SuTI0[OO 9Yy} YITA poyVTOOSsB 9DUBISGNS SNOUTSAI 10 A[TO UB O} ONp A[qQBqoid SI 9[qB} STY] UL pepiodss SUOTINIOS Jo soUBIvaddB APNOO 3UL« peztojooep =| Apnoja q009 ON qyooyo ON| ATI USITS 30299 ON yooye ON| ATI USIIS qoaye ON qOQH9ON| WayoON| 4apoon|* * * **umusy : qyourstp Apnoyo uUMOIG ; qoayo ON qooaye ON YSIUMOIg, iB = A ‘userid 0ing| A[7USI[Ss\-ys ies aiy qooye ON) +=WayYsaON| aye ON) “UONN[Os ouTpoyT . Tits) ) UMOIG peep ud013 sursusyo uUMOIG “a oa 13 uMoIg : * "mane -YSTUdIIIH| -MOT[OA O[Vq\- YSTUMOIG| ‘UAMOIG-U,1D|- YSTMO[ [OR orb) Ret 8 OS aS FUSS 2D UMOI_ |W9913-9AT[O OLLIaj-O1moulNay . 5) Apnoyo Apnojo ‘us01 cari Spnoypo qyooyo ON| ‘uUd0IZ a[Vq| ‘UdeLd OBI qooye ON Apnolp Apnojp|-uMoiq aed Apnopo £ [3G 3TIS fpaog) = '45 5. 2° umry Apnoyo u9e013 janmydins pue qoaye ON| PoIzofoood| UMOIG JUTB|| PeZ}IO[ODIq| pdzT1O[Odeq]| pezIojooed|‘uMOIq s[Bq -YSTNIG WYSTT Apno[p £pnojpp su wWRIpos Ooo ON) UMOIG JUTBA a9 ON aye ON 0ay79 ON Wao ON 309899 ON P2BION) WPayeON| Wage ON s}LjIU UNpes : p9z110 ‘O09 - * * * Oprxo q02aYo ON|-[00ep A[jIvg} pozptojooeqd|-9 p ANYstTS qo0y2 ON q002 ON qoaye ON y9y2 ON| YageoN| aye oN}-zed uasompséq pezt1o pezt1o[oo ava[o Jao ON|-[O09p AIBA) pozptojooed|-9 p ANYSIS 0289 ON yoaye ON 0909 ON #0999 ON) +=yOayo ON|SULBULsy)* * “pre omen poztio pazp10joo ree] 4009 ON|-jooop APB ojoo9q |-29 p AYYSTS qoayo ON ayo ON qoayo ON q0ay2 ON| + 4yoayge ON|SU gE U2 Y] * “prow oyforypes Apnolo | woer3-dAT[oO Apnoyo Apnoyo 4 pnop UMOIG-YST ‘uoo0is | Surmood ‘uosois [44518 'U 00818 *‘uaeaid |* *** oyeqd -U9013 O[Vd)-YSTMO[[AA|-oq ‘ud0I1H| Wolds oeg “UST AN OTLOX pearre 34arT -YSTUMOIG) UMOIG I43r]T UMOIg ofBg|-OATIO WBF] |-Ms SNOIIeT nolo Lo uMOIq u9013 uses =| fu re ry ~Vysrs i use13 maald -YSTUOIIH|-YSTMO[[A A) Ud9aId-OAT[O|- WSTUMOATG|-YST MO, [aA|-9ATLO BUSIT|-YSTUMoIg Ways ON UMOIG |-BATTO IGSr]| “ep; yO OLA ugo13-Y u9013 w001d u9013 uses | pours [> * > > ayeud -MOT[OA FYSIT|- YS TMOT[O Al- UST M O[[OA|- USTMOTLO X qoaye ON | Wea A YSIS qoay2 ON 70ay2 ON! -ystMoyeA “Wau “OD “O)-soyd UNTpoes u00.18-4 usa1d ud013 u90813 weal? pegs j-* °°: ep -MOTIAA JUST |- iat, ol -YSIMOT[OA|- USTMOTLO A 0999 ON wooid ANQSIS yoaye ON ay? ON) -ystMore A -“WazUT “9 “O}-efo wopsEjOg m9018-t Supuro.eq u9013 u9013 used peuys “MOTTO FYS!T| “MooL5-Ys,[OA|- YSTMOT[OA|- U8 iad See qoaye ON | UaeIs ANYSIIS qaye ON }0AY2 ON) -ySpAosye A -WazUy $°9°O) * * SHIERIV ' Apnoyo Apnoyo Angas Apnoyo |esuoqzut ssoy uaoi3- } qOaH9 ON) ATAU SILS oa "ia tay £14 UST [8/10/00 ‘Apno[p|-uaorq 4 Apnojp yoape ON Spnoj)) * spe opuelp Apnoyo Apnoyo Apyays losuoq UT sso] Spnop ay ON| ALYY SILLS) ‘pozopoooqg puw 19, Vd ApnoyO sooo ‘Apnoro}‘aaorq oped £pnojp Spnoy) e Apnojd) * spew reeuyK : 73/010 fisund Pi a advqqn2 | aGogqno POOMON 62! monox “op | monag “zg \POPMOUKD “HT| dnosopmag *L) MPOMOT “9! yunyy “gg | POMS “SF | 44D TI) vena “gg ‘IAHAOKOUHD "IAHAOYOTHD TAHAOLIG “SHONVLISANS AOIOD GILSVIg dO NOILYNINVXY a 269 KRAEMER—NATURE OF COLOR IN PLANTS. 1904] qoaye ON Ay? ON q022 ON qoaye ON yoayo ON qoayo ON qoaye ON qooye ON ADO ONT] 22 F-* uyUUB, sousoser0n uwd9013 YITM MOT Jaye ON) = ayja ON|-[e4 Jo ‘901s oINg| UWo9IId qUTB qoaye ON qoayo ON 4oaye ON yee ON qoayo On| * * UOTNTOS oUTpOT pnop WAOIG-US} TMOIG uMorq- UMOIG-YST Moped |‘ MOTTA | UMomqeyss | ott uns “uses 14 STT yoayo ON q0aye ON|-YST MOT [9 A|-MOTPOA 1 |S [T|-ud0ts doo d/-ystUMOIgG-QSTuUddIIH|-MOT[OA Yared) Odaj-OTUO MU Apnoyo yoaya ON qooye ON Apnoyo Apysyg| peztolooeq qooye ON} uweer8 a1¥q qoaye On| ATVUSILS yoaya onl °° umty soulsosa10N * “prow opmydyns pazoyooeq} pezyofoveq{ ysmn[q “paztopooeq| pez|1oTooeq qoaye ON} UMOIq JST] pazttojooeq| pezojooaq qoayo oN | PUB oyLyTU UINTpoOg Apnoyo 02999 ON yaya ON qooye ON| pozttopooeq, qoaye ON qoayo ON qooye ON! ATI MSITS| MOTTA opeg| * * ‘O9T4}TU UINTpPOg q0aqye ON| = Waa ON Wage ON} = waya ON qoaye ON} «=. QO ON! ~=—s ooo ON) «=—s AYO ON| © MOTIOA aT PPrxosod USorpAH qooge ON} =. ao ON yoaye ON yoaye ON yoaye ON qoayo ON qoaye ON qoayo ON qoayo ON|* °° * prow oe aye ON qoaye ON yaya ON yaya ON qoaye ON qo2y9 ON| =. aa ON qoayo ON qooge ON] * * * plow or Aoryeg u9013 Leos yh 3 MOL -YStMoTos m001d -u9013 sar joaye ON qoayo ON} = uW9ed3 a[Bq| - [9 A-ystueeTH} UMOIq YIBG] UodLd yUIRT ore: EOMOT) -OATLO o[eq| * 9BYd[NS sno UMOIG 04 —— woos WMOIG Moras MOTs | SupsuByo : “UdaId 4 WH ON) UMOIG IGS ATOA|-YST MOTTO A-USTMOT[[A A) UMoIq yIVC|-YsTUIIIH|-YSTUaerH) “Wetd-oAT[O] © * OPHO[Yo oVLiog moors Morfad uae43- poyysuoquy yaya ON qoaye ON} yYstueets AYSIIS}} UT By ATOA|-YstudeIs YYSVT| MoT[o4 oIng]|-Morpas ee qoaye on ANUS “o ‘| aeqdsoyd wmnrpog u9013 MOT uae18- poylsuequr _ yoaye ON qoaye ON| ystuser3 Ap) vuUTBy AIA! -[a4-YstuseID| Moped aing|-MoT[os cork qoayo on|ADYSTS ‘0 "O| ApruBsD UINISsBIOT uael3 MOT u9013- poyisuoqut yooye ON qoaye ON) ystucers ANYstg}} ur By AOA] -[oA-ystuseID| Moro oung}-MmoT[as foo Be qoayo on|Sq3ts ‘0 O)" SOT[BATV RUGOTS & B) Apnoya ATIU3TIS qoaye ON} = Apnoya Angas Apno[o 30999 ON yooyo ON! £14 a3 g| 413 13 ria qooyo ON) * * * Splov oyUBs1Q SGROES Apnoyo : ATVUSTIS yay ON Apno[p _Apnop 009 ON qyoaye ON] ATYYSITS Apnorp yooayo ON| © * * SpfoB [BIOTA sdry ndynd-ay} . : pad pad ; , avon DAL “8L|-w-yooe og | = OPO TE | OMMAL “Sh | sounig woe | wows, “egg | UML ‘9b juonspuDd “LI|pIobMmy “SL TIAHdAOKOUHD ‘PInuUuUOD —SAINVISANG AOTOD GILSVIg dO NOILLVNINVXY ‘T ~ ne aD?" I ne re Oe: ne KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, 270 “UOTUIOS ITOYOo[E 0} 1ayvAK JO UOTIPps Uo LpnoyD s qyoaye ON) | qoaye ON qyoaya ON qoaye oN yapaon)?-* tcf ct ttt uruuey, yoyo ON yay ON yoayo ON q0ay9 ON 2200 ON 8 ee WONNIOS suIpoy uUMOIG uMOIg UMOIG-YsTUdeIH|-YSfUseIs 0} YSTUIEI+ UMOIG-dAT[O dood | UMOI-YsTMOTIOA FUSPT|-ystMor[as IO pouayrBq ~ * * * * * * Uns dELiaj-OpMOuIMTy qyooye ON qoayo ON qoaye ON qoaya ON qoapgaons ccc ttt tte ocr aun aan Sapa pres 40299 ON UMOIG IYSPT 30299 ON Apnojp Mores aed -qns 4q pemorjoy oyu yaad yoyo ON UMOIG 9[Bd yooye ON qoaye ON qo0me ONE f 88 FR 32 Se aiuy umrpog qoayyo ON qoaye ON qooyo ON qoayo ON | Siete a etre eprxored uasompséR qyooyo ON qooyo ON qoaye ON qooyjo ON anges Gigi o oa es = Soe fais pre are) qooyo ON qoaye ON qooye ON qoaya ON qoagaon|" ct pre on sores UMOIG IYSIT UMOIG-YSTUIeLS FYSTT 9013 O[Vg| UMOIG-YSTU90I3 JYSTT\0} W9018-9AT]IO 74 St'T qoagaoni* * ttt ayeydyns snowag U9dIS-9ATTO UAMOIG IIIT uMOIG UMOIG-YsTU9eIS YYSyT|O} SupSuvyo ‘u9d1H| UMOIG-YstUdeI3 YIVq\0} ude19-sayJo 3 4 F1'T|-ystuser9 Jo pousytVg ~~ ~*~ apHorye onueg —-qoayo ON MOTIOA OBA Apystis pousyreg MoTo4-YsTUAeI qoayaoni> - yeydsoyd amrpos 4ooyo ON MOT[O4-YstudeI Apysts pouoyreqg MoT[o4-YsTUVeI qoqge ONI* *-*° °°" apruefo umsstiog UMOIE AOT[oA-YsTUIeT Ap ysis pousyred AOT[a4-Ystusery qoagooni* "°° ccc ttt SOeYIY pozt10[ooop ANYSIS yoayo ON qoaye ON Apnojo qoageoni** * "ttt sproe orueduo MOTOS FYSTT qoaye ON qoayo ON A£pnojo qoagoon)* °° ttt tte sprow peru MOH ‘OL «(dnayoneg) addy 7 wanoyog “6, danoyfog “98 uosfog “08 ‘LS ‘SHONVLSANS YOTOD ALVIGANYALN] dO NOLLYNINVXY ‘Il KRAEMER—NATURE OF COLOR IN PLANTS. 271 1904.] qooyo ON) por-ystiding qoayo ON yoayo ON 009 ON qoeyo ON pousppey qoaye ON qyooyo On| sete > Ses uae, por * * * eprpor uns pozt0[o00q onid qwoaye ON qyooyo ON qoaye ON ooo ON Me oppo ond qoaye ON qooyo ON|-SBjod pus oUuTpoy -Ls-YStay uUMOIq 04 por - uAMOIq uUMOIq SaySuen0 fi pr Dt! ie) SERS Ee eae chilebe ema Me wine SupSuvyo ‘ysyding|-yst MOTTO A|-|STueeiH| woerd-oaAy[O| woerl#-0AT[O|- VST p poe y\-dva ‘oniq dood |-Ystuaeiy) UWoeds-9AT[O/0 T1105 - opuoWMLY por eniq aye oer qOUNSTP ‘On{q Ing |-YS! MOTTO X yoaye ON qoaye ON qooye ON! poztofooeqd ays ‘entq-Ayg a -pak AMYSS}* * * ‘i ‘ Po : oar. Tt pozt1o[ooep por MOT po- por $8a]10 ~10[000p Sut UMOIG-YSt |-[NS YIM poAoyT qsoumys ‘per yuyeg|-US} MOT[O A|-[94 WoploH -MoTTed FU YSTPPot [VA |-YSTMOT[AA|-[00 Woy} ‘poy|-woseG ‘poyY|-MOT[OA [Bq )-[OJ ‘oILNTU UMTpog ystuooid pezT10foood Ges qsydind oyeg) WeeIs qUSPT|ATIUS TLS) yoyo on uoel3 OINd meetH q0qyo ON) * * O}fI}TU UMTpOg ooo ON|-GSIMOLLOA qoaye ON qooyo ON 4009 ON qoayo ON poyisusjzutr “dO per yqSIIS qooyo ON oprxosod uas01pAR por poy . por-Ust ; a port JUTBA|-UST MOTLAA yooye ON yooyo ON 09Y9 ON|-ISU9zU! “O *O POYISUIJUT ‘9 "O POr ISIS |-MOT[OA FUYBT) * * * * Plow oTTTBVH por pouopper pot-st por jupeg|-UStMOT[OXA| Favyo on AL 4 YU SITS YSIpper oped yoy ON PoyIsuoyzUl *d *O por quyey|-MOT[OA JULBT) © * * Pow oT[AOTTRS uUMOIG otdand u0013 en[q eng} per-ystding|- Fst p poy U9013-dATIO -USTUMOIG a a oniq deeq| weea3-ystntq|-9 AT[O yYsrT| * oyeYd[ns snor9,7 UMOI 0} 2uySueqo u9013 u9013 Oo} UMOIG ‘ant q-yst{ding| por-ystding|-YstuMOIg|-OAT[Odoog| Uoots-0ATIO/-YSt[ding) on[q osuojuy U9dIS-IAT[O| UWe0IS-9AT[O} * * opMoryo og enTq u9013 w0013 : usery|- YS t~daind) - Usd) -YSIMOT[O A -GSIMOL[IOA biten ea) uaety user} Uae1s o[vq| oyeydsoyd umrpog uoerd | u9018 0013 weer ani dMNg|-YSTU MOI -YSTMO[[O X-YSTMOTLIOX 015 wer weedy uded15) opruvsd UINIssB}0g 190.189-YsTMOT[OA 9013 we013 w9a13 ud018-YST 100.18-YSt 0} Sursuvyo ‘user ONn[q 9INg|-YSTUM OIG -YSTIMO[[O X -YSIMO[[O A\-UMOIG JYUSLT|-UMOIG 0) UVeTD user Wels SOT[VALV pot por po.- per-Ysty pot-yst : por dsop omg|-yst MOT[O A\-dand ‘sna -MOT[OAL MST -dind o[¥qd pel-joforA|) porystding poloamg)-MOTjoA a[Bd} * * *Sprov’ oruBsIO por ‘por-UstL por-Yst por pot por pot pol-Yst por deop orng|-yst MOTTO A\-daind doo -Mopjod USIT- UST [ding you osuozuy yor osuozuTdoop oing}-MoOpod ofed| * * * Spfow [eIOUTW ; en[q PF QOTOTA 0} [dmg] etdind deoq; oniq FIV) ong yysyy, onfq wsry)-ysttding oqdind) = onyq-49[OTA ond) * * * Iopoo penyeN . uo suanoy DIU0} F an: anjq amq-S- 5 TK — | non enum | mopoyy iz | -enogy” sq | "MPM “T9 |eypuroongy gg] ‘houvg ie | ‘wIOA. ‘ep PPPH “eK “SAONVISANG YOTOD dVS-T1T4D ANTG dO NOILVNINVXY ‘JII KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, 272 q0oyJ2 ON yuyd yurey q0ay9 ON yao ON q0ay2 ON qoaye ON per snugns| = Fate praghh tea uruUBy, 02Y2 ON} * 0EaHe ON 42ayo ON yaya ON q02Y2 ON qoaye ON 30aH2 ON} - * * WoRNIos Qurpor U9018-9AT1O eMC OTION. “YsTUMOIG deag) UMOIG-YsTUIeIH bas i capes 2x mee uaorg-ystiding oueJ-O;UOM WY q09H9 ONO} Sulsuvyo ‘poy qoayo ON qoaye ON |-ystidind yureg|‘por-ysttding PAHOA) 25°) = Ses + mnry TWALOIG asuBI0 lieieiGe ner por-ystdang Pad-YsTMOTIA A MOTI B[V |-YSTMOT[OA B[BA/IO PoI-YSTMOTJEA! PII-YSTMOT[PA, por dsapamg)-foy ‘sie wMIpes uUMOIG OTB q00J9 ON qoayo ON q0aY9 ON|Apno[o Suymooag q0ay2 ON 30aQy2 ON) * * * ayEITU UMIpOS 0aJo ON pow qoayea ON 1099 ON 4002 ON yaya ON pad aing| ‘aprxozed uasoup.iéq “uy Any at MOO quid AN ysts qoayo ON qoayo ON q0ayo ON qoaye ON pelamg)* ** >> Poe o7eD “uy ATU, ns m *O Uspyutd 009 ON vane " qaaye on q0ay9 ON pazammg|~* * ** prow orsorpes ystdind oped adimq| user3-dAtjo yurey -woers oped AOA per-ystding atdind y1eq wats * * apuydms snouag uMOIG u9013 uMOIG 0} Jursuveyqa -YStudois ole uMoId-GepaeeI)|-y sy UM Org eed): 4 FL 59029 ~ oa pol-ystiding oidind y1eq|‘per-ystpding) > -* apoyo omag uoeds YST| Wo9sd-ysntq IST BarRUe HO eoyaaet) “YSTMOT Ty aed uaeIp waaIy)| wath) - aeydsoyd umrpos 90418 ong u9013-YstInig Rocio arto 19918-YSTAOT]I A user ude13 asua}U] weay) -aprueso urnrsse}Og “ vageaiet ih oe u9013 ON UddLS-YSIN[ {| SuLsuByo ‘YstuseiH) WoIl3-YSLMoT[OA WeeIH) wudeTs osua}UT wuealpi-* +++: SHTRNTY poa-yst[dang per ong} Apoyo yByMouLog WOUTTVS JUTBT per 1q Sry -ysydind 4 ast Ti perdsop ama! SPIov NUE por-ystding per and) Apnoyo yey AMoutog UOUTTBS UTR T pew perdseq) pordeap amg - spe pEIBUTT ystidang ystidang eiding ystding por-ysttding per-ystding par-jaf0r \| 207OO TeNyEY : CHOMPT “9 uncom yong 1s] MT 1 | -oaus nosgodsyy -vge| -quo qungy “1p | -quo guts “op |-opeeioor "SHINVISENG YOIOD dVS-11aDQ AMdYNg JO NOILYNINVXY “Al 273 KRAEMER—NATURE OF COLOR IN PLANTS. 1904:] qooHe ON) YS peepuat'o"O 88910100 . Surmoooq ‘pew qooyo ON w098.Ld pol UMOIq - YS TN ][ A ysTUMord - ystdang qyooye ON qoayo ON goudosa10ny Sur por daaq)-sop, ‘qidand = yurey yoyo on) qoayo ON qoayo ON qooye ON qoqyo on) = edad Apastis pod qoayo ON ona yao ON uwo018-dAT{o dod uMorq-ystdang MoT[OA-YSTUDELH qoayo ON AMOTIOA-YSTUIOTH) UWAMOIG-YSTMOT[O A UMOIG-YSTMOT -oXk 0} Bu a} ; UMOIG “199 TS-YSIMOT[OA |-USTMOTLAL O[Bq . gous0so10N por yuyey | Susor ‘ot dind doudoso1ONY poi doaq Ssusoy ‘a~dind por-ystdamd yysry por-ystdang ia) . wnrupiab PLA “So aaynq “PLO qooyo ON qooyo ON qooye ON qooyo ON qOOoOON|* °° °° “UyuUByL yoyo ON qooyo ON qoayo ON yoyo ON qoqya ON} = *_ UOT NTOS duTpoy Ud0.13-9ATLO uMorq ofeg| UMOIq-YysTTding ud0.13-9A TIO o[dand doa ole J - OF ower Vv yooayo ON JOpOTA yuyeg -ystdand qo10LA yoayo ON qoogoa ON} * 7 55! gies po.-ysTuMoiq -inydyns Aq poMoy MOTOA-YSTUMOI OF YSTMOTPOA OTBq) — POL-YSEMOT[OA — UMOTg pos-YstUMOIg|-[OJ ‘o}LQTU TUMTpoS ysTuoo.s IST qooyo ON qooyo ON} UMOIq-YsT[dang|-uT 10] nO oben * * * oyLayTa urnrpog qooyo onl qooyo ON qoayja ON qoayo ON qoaye ON| oprxoted uasorpAHL Pp, uappoes APS quid AYYSIS| PoyIsuozUt “od “O q2H9 ON yoayo ON} * * * * * plow aT [BH p.ueppor AYYSIIS qoayo ON) Poy!suozuy *d "O wayo ON qooya ON| © * ** Plow or[Aorpes Ud018-9ATIO ony IYSTT ajdamg} UMorq-ystding 6 nt * 9 Byd[ns snoI19,T UMOIQ oO Sursureyo U99013-OATIO -YSTU9IIsS YOST ajdand-oasoy| UMOIg-YSTUITD |‘UAMOIG-YS TT dand| *~* opMoryo oft, Wd9aLS-YSTMOT[O A uders JST UooLs-YSINT JO[OIA) poesunyo Apysrs| * oyeydsoyd umtpos 10015-YSTMOTIA A usets Jysry) weeds osuoqUT oe. TIAL ~_ ajdand-ystumorg) *apruvso umnissejog oO} Sursuvyo TOD.LS-YSTMOTIAA u9018 IST wooi8 osuaquy)|‘ 49013 aang! opdimd-ystumorg)* * * °° © SONBATV poLYSTMoTOA| por-ysydand yysrT| — pos-asor yySry| Pep SUPT °9 *O porysyding| °° sppoe oruesi9 POL-TSTMOTIO perysyding| | — por-asoy) peyIsuoquy “0 °O porysydimg| * * * splow perour atdand-ystppoy por-ystding perysydimg} erdand-ysmmygq| —“oyd.md-ysyppaa| * IO[OO [BINVN vobuvsphy 1, | drums, “xy | 280997 poy “8| puosuoy 69 | TOWPME “V9 ‘panulju0) —SAINVISHAG YOIOD dVS-11a) AIdYNg 40 NOMLVNINVXY “AT PRINTED AuG. 6, 1904. PROC, AMER, PHILOS. SOC. XLIII. 177. R. [April 8, KRAEMER—NATURE OF COLOR IN PLANTS. 274 qoaye ON pays ON ; qoaye ON qoaye ON por-ystding per facing pets ulUuURyL dale ON) ¢ -YSTMOT[A A qoaye ON qoay? ON qoaya ON|Surai00eq ‘Pp 2 A) * * WONN[OS sUTpPOT qyoayje ON pos-YstTMojl wy re qprrdand uMOIG Pratl oo Ct pees umnye -ystuoe1s 918d 19018-dAT[O FIV par deed) per-ysTuaord uMoig|Je[01A daaqjuiy-oruomury 900d -saiony ystuse.s e YIM UMOIg -ystppat 0} Sut haw quid qureg|-SuByo' Moped a[Bd) POT USTMOTION parystmorax) pargstdind perored| © “© ~ *-" - am peg “pre dL por |-suoyur pus [Os YIM Psmoy por aed por-YStMOTIA A |-OSUBIO daa q| pot0yset nM a8 YWSr]| per-asuvs0 dasqd}-jo} “aI WENTpOS pezt j erieg -qst qoayje ON |-1O[099P ADUSTLS qoaya ON|-ep ALIIG4ELS -UMOIg 7 watt 0092 ON|* * a3ETU UiNTpeS qoayo ON| = Pay stay “Oo i bags *2 ‘Q| peytsuasat "9 *O -ysyjdind 3431) PAF-YSEMOTIO A aprxored uasoupsH quid yuyeq) — peyysaoyuy"o oe sc -g‘olpagisuayay“9°o| © parysdimd) PAF USTMOTTON) °° © PPR ITTED yurd jureg| = peyysueyur’o “O payisuezut 9 "| peyisueqat ‘9°C) «= perystidind) = PAF USEMOTIPA > + + pre afores uUMOIG -OUTM 0} see nas : -ysyjdand eT’ qov[q-ysin(a|-suvgo staat par-ystimorg |-ysrumoig ded — ee, aryeydyns sno "MOT OI, OF -u9ei3 oO} Sut | yak - nee oviq-Ystaoaty) por daog) par-ystumorg UA 1q-USTl -Suvyo “ysyding apyorye WRT MOTI. -ystue013 ied usety weer) Wears WYSE uder3 34ST qaage oN “eyeqdsogd umIpes MoTos -ysTudei3 FUSTT user users) 9013 SATIO 19014) WIdTS-YSEMOTPS A -apraeds mIMSSFIOg UMOIG-YS, pat muMaIg ’ moypos uMOoIq |UMOIG 0} Sur |O4 suyeueqo oo Sursueqe dec -ysyu0a13 JUSVT|OF Sursuvyo a etal ‘aarp ‘u9eI3-9 ATL O waads)) ‘W8ars-YSTMOTPO A) ~ ~~ Suey pos oped) por-ystidand sols acct fog poysuaquy’o°O| = parystidind parasuwo -™ + * *Sppe ormEmuQ por o[8d por asOy -95UBIO daaq| per-qstsoreA por-ystiding par-asueso dead) © * “Sppe PRDUT Utd pew pew par dood ystpped| par doap imal + > ZO7OO TRANIEN fea Pe SS wrt waypey ‘F asoy “cS yrurvhyy “og | woHDULDD “OL dyny “oF | -upsad aenozy = qoaye ON qoayo ON qoaye ON qoayo ON uMOIG, enjq deed ~~ 7 : Apnop qoaye ON|‘P 2 t- ystdand UMOIG Apnoya -Y STMO [2 Aj ‘POLUSTMOTION uUMOIG Motos -YSIMOTTAA -ystuset3 JST tg qoaye ON|-MOTIOA = FUTBT qoayo ON| Pet-YSTMOT[OA qooyjo ON| Per-YSTMOTIOA UMOIG-YSTPpor anjq dead pos-YsTUMOLE 19015-9ATTO ad ud -md - ystumorg|-YS TM OTLA A aid | UMOI 0} oUt -and - ystumorg|-sueqo teen eg) od -imd - ysTumorg u9e818 FIV ; rao asuvyo ON) ‘por-YSTMoT[o A Apnoyo asuvyo ON) ‘por-YSTMOTIOA por dood pow suanoy ALtaqQuoly “Sl | ado ‘8% "SHONVLSANS AOTOD avS-TTA9 GIy 3O NOLLYNINVXY “A eet: KRAEMER—NATURE OF COLOR IN PLANTS. 275 1904. ] qOayo ON pope ON qooge ON qo0y9 ON qoayo ON qoayo ON Me ON oe tH SF UIUUB YT, 7 1099 ON qoaye ON qaye ON 30959 ON 10a ON 1009 ON qoaye ON} "+ * WONNIOS ouTpoT uUMOIG “YU STM e7 18 “YUsStuoois ae W0913-9ATIO UMOIG YIVQ|IO MOTT dood UMOI OTB por-YsTuMOorg: UMOIG)| * UIN[B dLL19J-oruouury 0299 ON qoaysa ON q0aH? ON|-Ystmorpos ane d pos-ystdang goaw ON] popysuojuy oO] es wnTy UMOIG IYSYT)L0 | por - PUsTUnoug qoayo ON per Ws MOTIAA OTB |-YSTppot aust Tl- -uspnorpes o1%d paAory ener mnpoR 40299 ON qoaya ON q0IYo ON 10999 ON pozt1ofo0aq, qoaye ON} =uMorq Apystys|* * * + * o7gTU TANTpog qoaye ON qoaye ON Jaye ON sitet on oO qoaye ON qo ON qoeye ON) * oprxored ueSorpéAy qoay9 ON qooygo ON qyooyo ON|-Ur A[3u01s 2°O qooygo ON qoaf9 ON sOGne ON]! Set rat ples oT [By 10959 ON qooyya ON aye ON eager — 3) q00J9 ON y9N9 on qOOQJOON|" * "°° prow or Aoryeg 10aH} ON uadis-YysTUMOIg|-ystusers = yqsrT|ATIEStTs ‘pay YSIppou|-YsIppet FUSIT) port-ystuMoig|* * * *‘ajzeydns sno.ws10,7 “Ystmoypak aT Bd usei8 daaq| uMOoIq-YsTUseTH ciieuene” bese UMOIq-YSIppoy| UWAMOIG-YsTppey) * ** @pHoryo opueg WddLS-YSIMOTIOA| WoIts-YSTMOT[OA YSTUIeIH| uay} ‘ONG JYSsg|-1ojooop YSOW[Y| WMOIq-YysTuseIyH etdind e[eq| * * * ayeqdsoyd umutpog ou A[}U9IB ved d sy MOT[OA-YSTUVOIH wedi ont W438 ud018 IYSYT| UWMOIq-YstudeIyH ystidang) * * * opruvdo umnissvj0g ~YSTMo[o£ Kr [¥d| AMorpe<-ystuserp uwaety weep UdeL3 YYSTT| WMOIG-ysTUseIH ysTuUMOIg| * * * * * SOTexTy woeeId peytsue4 pogis PoyIstoy “YSpMoT[eA 9 | B J |-UT JV MOOS “Dd “O q0aH9 ON)-UazU] YonU “d ‘O) peylsuajut dO) peyrsuazur ‘do ‘O-ur Apysys “9 “O° _“‘Splov orUBsIQ ou kyuoreddy PoYIsuazuy *d *O q00H9 ON| = od-YSTMOTIOX -u9qUT | DO pey]suayuy ‘dD ‘O) poyisuazuy a “O|* * * °° Splow [RIoUT pouepper ANYSIIg poy por doaq per qqsug) per 0} ysTyut por Lamy 07 POF-USTMOTOA fete eee IO[OO [BIN}VN addvoug ‘ge |addvunpog 1| wg -s9 umpoy +e | qqnyar “14 '9¢| s uowojog “1g | AHeqmoug -6¢ ‘Panuyjuo)—SAINVISANG AOTOD dvs-T1aDQ aay ce fe ee ae 40 NOILVNIANVXY ‘A KRAEMER—NATURE OF COLOR IN PLANTS. [April 8, 276 5 per yoay9 ON yoyo ON 0299 ON 402099 ON 409H9 ON qoaye ON|-ystidind yureg)* * * 7 7 7 7 uuueL uUMOIG -ystidind 0 qoayo ON 4999 ON qooye ON qooyo ON|SursuBvyo ‘user qoaya ON sae Ont 22 UWOTNIOS sUIpO]T u9013-9ATTO 9018 yred uses deaq| us013-ystuMoiIg UMOIG-YsIn[g| wose1s-ystydimg W9d013-9AT]O| * ~ * UIM][R JLLIajJ-O1m0mla yy ~ weeds uUMOIG per0joo qoayo ON|-YSTMOT[OA 9 [Bd |-YSIMOTTOA AVA yoayo ON JO[OTA 10 YSTy qoaya ON| UMOIG-GSIPPey|" ~~ * " ~ " * * wny past UMOIG 0} SUT - * * prow opmydyns girs por-YstMoya x |-Ysrumoig daa q)-suvyo “ysTpped qoqyo ON) UMOIG-YSTPpoey) AOj[oA- ySTU MOTT Baa bites Bice ‘ju WnIpoSs udeds AVYSIIS _ yooye ON| ees ANUS qyooye ON yoaye ON yoaye ON qoaye ON | gaan aes LU uMITpeEs “pol per 4yooyo ON qooyjo ON |-YSTMoTeA Jurey qoaye ON|-esue10 Apystys qoaye ON qwaye ON) oe eprxozed uasorpiQ por par yoyo ON qooye ON |-YSTMo[os — qoaqye ON |-o3uvI0 ser aye ON por oma Ste Rep eg PHB 9TTBD qooyo ON qooyo ON|-YSTMoT[OA qurey qoaye ON|-e3uvI0 ANqaiys qoaye ON per ee oie teres Mery pro arpforyes uoo1d use13 -o9SuUBIO UTR] udeis YB use13 doaq| Ws9eIS-0AT[O O[¥d aniq OSrpuy|-ystumMoiq ¥IVG eniq daag)* ~***~ aeydins snoLmag . uUMOIG oO} Sursuvyo ystjdaind 03 waeai3 19d18-9ATIO m901s YB use13 deed)‘ a9 013-9ATLO SULsuByo ‘YsSmM[_|-YstuMoig ¥ IBC uaelIs-OANJO)" ~~ ~ °°” apLoTyD oa poeyisueyur UMOIG MOTD. 90.19-YSTMOT[OA| Wo91S-YSTMOTIO X use13 osuoquy|A [9 UY STS “9 O-YSTUs0I13 O[Bg u9ads)|-Y STUIIIS apeg) > ~~ ~~ ayeydsoyd umrpos U9013-YSTUMOIG| WOdTS-YSTMOT[O A U9013 VSUd}JUT| PoYyTsudjUT ‘9 'O| UMOIG-YsSTUVeIyH weer uaais ated coe. ae eae eprueio MESO { MOTT? 00.13-YSt -ystuaeis oO} YsTuser)| 00018-YsTMOT[O A! -MOT[eEA osuezUT| poyIsmoezUT 2:0 UMOIg-YsTUVeITy uae Surdueyo WA ote? ete > SHTeNT¥ : por MOTI. per pouopper AYA qooyo ON|-YSTMoT[os 9 at . d ON S L , d par-aS UBIO |-YsTMoy[od warT qojoo asoyj~ ~~ Spe oes por-YSTMOTOA] YStppor AYSS|-ysdand 4 y 3 LT|-ysyaMorq 9 | Vd pet-osuVIO) por-YSTAOTOA zojoo asoy; * * ~~ * ~*~ SpHpe [SUT pod-ystuooty) pot-YysTuoody por-YsTuood pot-YSTUMOTE poi dood por-ystding por-ysquaap) © © ~~ * > 20709 PEINIEN ooyoM °0S 490d =“sT “APA “TS Auwaqyong = *9 40 “OS yunyy “th soy “SS ‘SATMIONING ONIMOTOD AVAT dO NOLLVNINVXO ‘TA 277 KRAEMER—NATURE OF COLOR IN PLANTS, i qoayo ON por-ystding qoaye ON qoaye ON qoaye ON TORR GT 82 Fe Sr Fe SES uyuURy, qooyo ON qooye ON qooyo ON qooyo ON qoayo o Wane ON. 2 9 eS uOTN][OS ou re W9013-9AI[O U90.13-9ATTO = atte sa “sie UMOIQ-YSTUIe.Ly uUMOIG-YSsT[ding |0} Jursuvyo ‘ontq dooq|o} Sursuvyo ‘ond oniq deeq udeI3-9ATIO|" * °°’ UWIN]S OLej-OTUOUULy MOTO. poztojovep Ayre por-ystding penwidnd qyoaye ON qooyo ON|- YSTUDOTS ySPT]* st tt wmv UMOIG |-USTMOTLO Pe Ee i ale bie. Sik harem em ce atk ee Sirsa GAMly aeees Steen mret Tal ic cpit, PG ot ily prov o1mydyns UMOIG JYSLT|-YSIMOT[OA FUT A|-Suvyo ‘pas-ystMoyjar Pot-YSTMOTIA A POL-YSTMOT[OA] UMOIG-YSTMOTIOA|GITM POMOTLOJ ‘LIU UINTpOg qooyo ON UMOIG JULB MOT[OA-YSTUMOLE qoayo ON qoayo ON ystuMOoIg)* *"* * "°° oyETU UINTpog poeztojooep AaB por-ystding qoaye ON yoaye ON qoayo ON qooye ON|" °° 5 eprxo1od uaso1psAH pezLiojooop Apaed pos-yst[q ind POL-YSTMOTION qoayo ON qoaye ISIS JOOS ON FS So Seton: plow OTe) per poztdopooep Ape por-ystding poL-YSTMOT[A A qoqye ON|-ystmorped ANYSIS WOO Of <8 er ea prov of Aoryeg UMOIG Y -YStUoorSopeg) WMorq-Ys~ppey aniq deed ut uornyos on[q deeq uaers-0AT[O| °° °° * ayvyd[ns sno119 7 uU U9ddIS-9ATIO |-OATIO 0} AlpIder 078} UMOIG-YsTUseTH| «—- UAL OIG-YSTppey OF SupSuvyo ‘ontq dooq|s urs u v yo ‘oni! -Wdrooad onjq dooq mweer8-0AT[Q|~ «°° * °° apMolyo dLI9 UMOIG ULB UddLS PST UMOI JUST) UWMOIq-YSTUeeTH) UseIS-9AT[O JySVT WMOTIOAL <<: 2° ot ¢ oyeydsoyd umnrpog ; uMorq IYI] A104 weer UMOIG pos-yst[ding | per-ystumoiq daaq YSIMOTJOA]" °° °° * opruesod UMISSBIOd UMOIq YSTp -pol ‘BILUOTI UB uUMOIg? ueolp uUMOIG UMOIG-YSTPpoy| ITA “Weetd-OATIO NIMES te a oe ats SOT[RATV por por poet poztiojooep Apavg| per-ystdand dood POL-YSIMOT[IA| -YStMoTos APYSIS|-ystmorpod AYSYS|-USIMoTpos 4 ut oe aay sitet SPlOV OTUBSIO _ pezttojooep Aprea per-ystdind deaq P2t-YSTMOT[AX PoL-YSIMOT[OA Por-YSTMOTJOA |-USTMOTTOA FUTV A) Splov [BIOUTV AMOT[[OA-YST UMOIG 0} YSU Ystppoy pol FrVq pol yrVd por yareq pow-ystueeIp|: -* ss. ct IO[OO [BINIVN “y202g “29 ane I 8 poombog "19 YO “9L doy “bh Mano yang “ph ‘panuyuojy—SAIdIONIXg ONIMOIND AVA 4O NOILVNINVXY ‘IA sin aelpgal di aa a 278 HEWETT—PRONOUNS IN ENGLISH LITERATURE. [april9, THE HISTORICAL USE OF THE RELATIVE PRONOUNS IN ENGLISH LITERATURE. BY PROFESSOR WATERMAN T. HEWETT, PH.D. (Read April 9, 1904.) In examining the manuscript of a new volume submitted for pub- - lication, I was struck with the fact that the relative pronoun which was not used by the author. The question arose, whether there was a portion of our country in which, through historical or possibly educational influence, the use of ¢ha¢ prevailed in place of which. In my subsequent reading, I marked the use of these pro- nouns in order to determine their literary use. Many of the characteristics of literary form depend upon the choice of the pro- noun adopted. The use of one or the other pronoun is a character- istic of the style of representative English writers and lends a special quality to their form and expression. The Germanic languages did not possess a distinctive relative pronoun. The place of such pronoun in Old English was supplied by sé, séo and fae, also by the indeclinable demonstrative form pé (the), which was frequently added to the article, and, though less frequently, by the interrogatives which and who. .What (hwaet) as a relative occurs first at the beginning of the thirteenth century. Following the Conquest, the use of é (the) as a relative declined, due, possibly, to the increasing tendency to use this particle in place of all the forms of the definite article. About 1200, the neuter faef was, in general, used as a relative in both numbers and in all persons and genders in the nominative and accusative cases. This use may have been promoted by the influence of the French conjunction gue. The interrogatives who and which were used, but only in isolated cases, as relatives, who referring mainly to persons and which to things. By the time of the translation of the King James version of the Bible, in 1611, the development in the use of the relative pronouns had attained certain distinct features. The most striking differentiation in use consisted in the fact that shat was made to refer to pronouns and wAich to nouns. The use of which had constantly increased and had gradually displaced that, and who and what had gained in frequency of use. The 1904.) HEWETT—PRONOUNS IN ENGLISH LITERATURE. 279 present tendency in literature is to employ who and which at the expense of the earlier ¢ha/. Every scholar will judge from his own use, or from the environ- ment in which his speech has been formed, in respect to the frequency and naturalness of the use of which and ¢hat in his own case. That which we do instinctively is the test of familiar expression. Writers upon the use of language-in rhetorics and popular grammars exhibit great diversity of judgment respecting the use of these pronouns. Dean Alford, in his book upon the Queen’s English, fourth edition, 1874, in speaking of the use of who and which, remarks: ‘‘ Now we do not commonly use either one or the other of these pronouns, but make the more convenient one ‘hat make duty for both. We do not say ‘The man who met me, nor the cattle which I saw grazing,’ but ‘The man ¢hat met me,.the cattle haz I saw grazing.’ ’’ Bain, in his Higher English Grammar, says that who and which are most commonly preferred for co-ordination, but that they may also be used as restrictives. ‘‘ However, shat is the proper restrictive, explicative or defining relative. It would be a clear gain to confine who and which to co-ordination and to reserve shat for the restrictive use alone. In the sentence ‘ His conduct sur- prised his English friends wo had not known him long,’ we mean either that his English friends generally were surprised (the relative being in this case co-ordinating), or that only a portion of them— namely, the particular portion that had not known him long—were surprised. The doubt would be removed by writing thus, ‘ His English friends ha¢ had not known him long.’ So, also, in the sentence ‘ The next winter which you will spend in town will give you opportunity to make a more prudent choice;’ this may either mean you will spend next winter in town or the next of the winters when you are to live in town, let that come when it may. In the former case which is the proper relative, and in the latter case that.’’ According to my own impression, the ambiguity in the sen- tence ‘* His English friends #a/ had not known him long’’ would not be removed, as the author thinks, by the substitution of ‘haf for which in this case. Genung, in Zhe Working Principles of Rhetoric, 1902, says: ‘«Typically, the relatives who and which assume that the ante- cedent is fully defined in sense, their office being to introduce additional information about it. They may accordingly be called 280 HEWETT—PRONOUNS IN ENGLISH LITERATURE. [April9, the additive relative, and are equivalent to a demonstrative with a conjunction, ‘and he,’ ‘and this,’ ‘and these.’ The relative that assumes that its antecedent is not yet fully defined, its office being to complete or restrict its meaning. It may accordingly be called the restrictive relative, and may generally be represented, by way of equivalent, by an adjectival or participial phrase.’’ Professor Hill, of Harvard, says: ‘‘ Few good authors-observe the rule that who or which should be confined to cases in which the rela- tive clause explains the meaning of the antecedent or adds something to it, and ¢ha¢ to cases in which the relative clause restricts the mean- ing of the antecedent. This rule, however helpful to clearness it may be in theory, few good authors observe; considerations of euphony prevent adoption of the rule’’ (Principles of Rhetoric, revised and enlarged, page 136). Meiklejohn, in his Auglish Language, says: ‘‘ Zhat is generally employed to limit, distinguish and define. ‘Thus we say ‘The house ¢hat I built is for sale.’ Here, the word ¢ha¢ is an adjective limiting or defining the noun house. Hence, it may be called the defining relative. Who or which introduces a new fact about the antecedent ; ‘hat only marks it off from the other nouns.”’ We thus have here representative opinions from English, Scotch and American scholars, who base their judgment mainly upon their practical experience of language and not upon an examination of the literary monuments in different periods. It is our purpose, therefore, to ascertain the historical use of these pronouns and to determine the frequency with which they occur in representative works in literature, since the period of Wiclif’s translation of the Bible. An examination of the two texts of Layamon’s Brut, issued about seventy years apart, show how complete the distinction between these pronouns had become in that period. In the older text (of about 1205) the earlier relatives of different genders as well as fe are used, while the later manuscript B. (of about 1275) represents these pronouns by a uniform fae? (that). A. (Line 13,827) An alle mine liue, pe ich, iluued habbe. B, In al mine liue, pat ich ileued habbe, In the century which follows, who and which occur, but less frequently. In the language of Chaucer (1340-1400), ¢ha/ is the 7+ eo 1904.) HEWETT—PRONOUNS IN ENGLISH LITERATURE. 281 prevailing relative; who, whose and whom occur but in few instances, and may then relate either to persons or things, as in Shakespeare. Chaucer stood more under French influence as regards language than his great contemporary, Wiclif (1324-1384), who in his translation of the Bible was influenced more by Latin constructions. If we examine the Morte d’ Arthur of Sir Thomas Malory (1400-1470), which lies intermediate in time between Wiclif and Tyndale, we find in 555 lines 30 cases of the use of that as a relative, 6 cases of the use of who or whom in indirect questions, or as an indefinite relative equal to whoever, while which (the whiche) occurs but once in the nominative and once, ‘‘ for the whiche,’’ governed by a preposition. This shows that sha? retained its supremacy in the fifteenth century. If we now compare the use of the relative pronouns in Wiclif’s (1384) and in Tyndale’s (1526) translations of the Gospels, which are separated by about a century and a half, we find the following results. The approximate number of times that the relative pronouns which, that and who occur in the four Gospels in the Wiclif and Tyndale versions is as follows : In Wiclif’s version of the Gospels which occurs 29 times in Matthew 18 « & Mark 97 * « Luke 27 “ “ John 171 “ the four Gospels In Tyndale’s version which occurs 135 times in Matthew 61 “« «« Mark 241 “« « Luke 125 “ “ John 562 “ « the four Gospels In Wiclif’s version that occurs 205 times in Matthew 84 « « Mark 284 « « Luke 228 “« <«* John 801 “ « the four Gospels 282 HEWETT—PRONOUNS IN ENGLISH LITERATURE. In Tyndale’s version that occurs 120 times in Matthew 78 “ 161 « 144 “ 503 “ In Wiclif "s version «© Mark “ Luke « John «« the four Gospels who occurs § times in Matthew 8 « In Tyndale’s version who occurs 13 times 10. 14 21 oe In Wiclif’s version whose occurs 1 ° 3 times 5 9 “oe In Tyndale’s version whose occurs O I 5 times 5 i Il ae «© Mark «© Luke «¢ John « the four Gospels in Matthew « Mark « Luke « John “* the four Gospels in Matthew “© Mark * Luke “ John ** the four Gospels in Matthew “ Mark “ Luke “ John “ the four Gospels [April 9, The relatives shat, which and who occur in Wiclif 1043 times, that 801 times or in about 76 per cent. of the cases, which 171 times or in 16.4 per cent., who or whose 71 times or in 6.8 per ot eee 1904.) | HEWETT—PRONOUNS IN ENGLISH LITERATURE. 2838 cent. of all cases. In Tyndale’s version we find a change, the same pronouns occur 11 50 times ; which has gained in frequency of use, occuring 562 times or in about 50 per cent. of all cases, shat 503 times or in 44 per cent., who in 85 cases or in about 7.4 per cent. To summarize: ‘hat occurs in Wiclif’s version in 76 per cent. of all cases, but in the Tyndale version in only 44 per cent. of such cases, while which, appearing in but 16.4 per cent. of such cases in Wiclif, has risen to 50 per cent. in Tyndale, and soon becomes the leading relative. In Tyndale’s translation of 1526, a usage was established which was preserved with only limited exceptions in the King James ver- sion of 1611. As religion appeals to the strongest convictions of our nature, and is associated with glowing feeling, the fixed forms in which truth is conveyed in the Bible have stamped themselves upon human thought and expression. From the restricted use of which in 1200 it had in the fourteenth century, the period of Wiclif and Chaucer, attained a recognized currency, while 150 years later (1526) it divided almost equally the sovereignty with ¢ha¢. The dominant use of wAéch with nouns is a fact which we might have anticipated from the primitive meaning of which, hwi-lic or hwa-lic, of what kind, how constituted, like the Latin gualis. Sub- stantives naturally possess character or quality, and the relative in referring to them means of which kind. That merely identifies and does not describe; similarly, who indicates usually an individual. Thus in Shakespeare, ‘‘I have known those which (gualis) walked in their sleep, who (equal to and yet they) died holily in their beds’’ (Macbeth, V, 1, 66). Quoted by Abbott, Shakespearean Grammar, page 182. Which is uniformly employed with proper names: ‘‘And thou, Capernaum, which art exalted unto heaven’’ (Matthew 11 : 23); ‘¢Then cometh he to a city of Samaria, which is called Sychar’’ (John 4: 5); ‘‘ For he was father-in-law to Caiaphas, which was the high priest that same year’’ (John 18 : 13); ‘‘ The same day came to him the Sadducees, which say that there is no such resurrection’”’ (Matthew 22 : 23); occurring in such use 151 times, while sa¢ is similarly used but 5 times. In Tyndale’s version of 1526, which refers in the Gospels to a noun about 418 times, /ka¢ to a noun 119 times, a total of 537 times, or in the proportion of 78 per cent. to 22 percent. Which 284 HEWETT—PRONOUNS IN ENGLISH LITERATURE. [April 9, refers to a noun denoting a place or thing 153 times, to a personal noun 265 times. Wich refers to a personal, indefinite or demon- strative pronoun 144 times. TZhaf¢ refers to a non-personal noun 77 times, to a noun denoting a person 42 times, or a total of 119. That refers to a pronoun 384 times. Out of 602 cases of the use of a simple relative referring to a pronoun, ‘haz is used in 64 per cent. of all cases, which in 23.5 per cent. of all cases. | The limited use of who in the Gospels in Tyndale’s version is shown by the fact that out of about 1165 cases of the use of the simple relative, who is used only 55 times or a little more than in 5 per cent. of the cases. The two translations of the Bible by Tyndale, 1526, and the King James version of 1611 present often kindred features in the use of words. The translators of the King James version adopted substantially the usage of the version of Tyndale. Nothing shows the dependence of the translators of the King James version upon Tyndale more than a comparison of the use and relative frequency of certain forms. We note a striking change which the language had undergone since the period of Wiclif. The relative pronoun which refers in the greatest number of cases to nouns, the relative pronoun “hat, in addition to its use with nouns, is used almost universally with personal and indefinite pronouns. The form of two petitions in the Lord’s Prayer illustrate this usage, and have remained fixed in liturgical service to the present time: ‘Our Father which art in Heaven,’’ ‘‘ forgive us our sins, for we also for- give everyone /ha/ is indebted to us.’’ The relative pronouns which and that occur in the four Gospels in the Tyndale version 1065 times. Of these, ‘haz is used 503 times,and which 562 times. The use of the relatives which and that in the King James version does not differ greatly from the use of these pronouns in the version of Tyndale. In Tyndale, the relative pronoun /Aa/ is used 32 times, where which is substituted in the King James version ; which takes the place of shat 4 times, and which is used 6 times instead of who, of the King James version, while in 60 cases an equivalent expression is used instead of a relative pronoun. In Shakespeare, if we take the AZerchant of Venice as represent- ing fairly the plays, /ia¢ is used 75 times, or in 83 per cent. of the restrictive clauses, while which is used in the same class of clauses 20 times, or in about 17 per cent.; /Aa? is used in co-ordin- ate clauses 11 times, or in 32 per cent., and wich is used 23 times, ~yeaar es 1904.] HEWETT—PRONOUNS IN ENGLISH LITERATURE. 285 or in 68 per cent. of such cases. The usage which we have found in the King James version, and earlier in the Tyndale version, occurs also in Shakespeare. In the above play, ‘haz refers to personal nouns 15 times, or in about 88 per cent. of the cases, while which refers to personal pronouns but twice, or 12 per cent. Who refers to personal pronouns 26 times, to nouns 8 times, to animals personified once. In the entire play, /ha¢ occurs 122 times, equal to 62.5 per cent., which 73 times, or 37.5 per cent. Which is used in restrictive clauses 20 times, in co-ordinate clauses 23 times. The usage of Shakespeare is thus very flexible, showing greater variety and greater freedom, as we should expect, than occurs in the version of the Scriptures. The relative pronoun was omitted in restrictive, but not in sub- ordinate clauses. Who originally referred to things as well as to pronouns, and such use is familiar in Shakespeare. Thus, in the Merchant of Venice, the Prince of Morocco, in describing the three caskets, says: ‘‘ The first of gold who (which) this inscription bears, who chooseth me shall gain what many men desire.’’ ‘‘ The second silver, which this promise carries, who chooses me shall get as much as he deserves,’’ A little later, ‘ha¢ occurs, often with great uniformity, apparently to lend smoothness to the verse. ‘‘ In the prologue of Fletcher’s Faithtul Shepherdess (1610), which was probably not written by Fletcher, which occurs, but ‘ha¢ appears uniformly in the remaining acts of the play’’ (Morris). A century later (1726), we find Swift using the relative ¢ha¢ when the antecedent is a pronoun, thus following the usage in Tyndale and in the King James version of the Bible. In the eighteenth century, there was a manifest effort on the part of certain writers to promote the use of who and which at the expense of ¢hat. We have in No. 78 of the Spectator, Steele’s humorous plea in behalf of the restoration of who and wich to their ancient rights: ‘‘ We are descended of ancient families, and kept up our dignity and honor many years, till the jack-sprat shat supplanted us. How often have we found ourselves slighted by the clergy in their pulpits and the lawyers at the bar. Nay, how often have we heard in one of the most polite and august assemblies in the universe, to our great mortification, these words, ‘ That that that noble lord urged’; which, if one of us had had justice done, 286 HEWETT—PRONOUNS IN ENGLISH LITERATURE, [April 9, would have sounded nobler thus, ‘ That which that noble lord urged.’ Senates themselves, the guardians of British liberty, have degraded us and preferred #ha¢ to us; and yet no decree was ever given against us. In the very acts of Parliament, in which the utmost right should be done to everybody, word and thing, we find ourselves often either not used, or used one instead of another. In the first and best prayer children are taught they learn:to misuse us. ‘Our Father wich art in Heaven’ should be ‘ Our Father who art in Heaven’; and evena convocation, after long debates, refused to consent to an alteration of it. In our general confession we say, ‘Spare thou them, O God, which confess their faults,’ which ought to be ‘who confess their faults.’ What hopes then have we of having justice done us, when the makers of our very prayers and laws, and the most learned in all faculties, seem to be in a confederacy against us, and our enemies themselves must be our Judges?”’ Steele’s view is specious, and is not based upon an accurate knowledge of the historical use of the relatives, or he may have had in mind a contemporary fashion in literature which he sought to counteract. If so, it is not clear against whom his shafts were directed. \ In the Sir Roger de Coverley papers in the Sfecta/or, written by Addison and Steele, the relatives which and that occur 531 times ; of these, which is used 353 times, ‘ha¢ 178 times. Which is used in restrictive clauses 179 times, or in 53 per cent. of all cases, shat 161 times, or in 47 per cent. of all cases. Which refers to nouns 255 times, ¢ha¢ to nouns 129 times. The influence of an anteced- ent modified by demonstrative or an indefinite pronoun, to which in certain instances the choice of the relative may be due, is shown by the fact that which refers to a noun so modified 83 times, equal to 76 per.cent. of such cases; ‘hat refers similarly to a noun so modified in 26 cases, equal to 24 per cent. of such cases. TZhat refers to a demonstrative or an indefinite pronoun 39 times, equal to 76% per cent. of such cases, which, 12 times, equal to 23% per cent. We see here a reyival or perpetuation of the usage of the earlier centuries. In spite of the great influence ascribed, apparently erroneously, to Addison in re-establishing the use of that, he uses this relative only one-third as often as which. In Macaulay's essay on Milton, the relative wich occurs 191 times, “laf 7 times, total 198 times. Which refers to noun ante- 1904.) HEWETIT—PRONOUNS IN ENGLISH LITERATURE. 287 cedents 174 times, or in 9g per cent. of all cases; shat refers toa noun antecedent but once. There is a striking use of who as a rel- ative. This pronoun occurs in all ror times; referring in 58 instances to a noun, and in 43 to a pronoun; to a personal pronoun 6 times, to a demonstrative or indefinite pronoun 37 times. This is the highest proportion obtained in the examination of any author. It shows a distinct mannerism, affecting noticeably the style of the historian. Zya¢ as a relative occurs only 7 times. Which is used in restrictive clauses 198 times, or in 97 per cent. of all cases ; “hat occurs in the same class of clauses 6 times, or in 3 per cent. of all cases. Which refers to an indefinite or demonstra- tive pronoun ro times, or 71 per cent.; shat 4 times, or 29 per cent. What is used 17 times. Which is used to introduce co- ordinate clauses 6 times, ‘haf in no instance. Waich refers to an indefinite or demonstrative pronoun 13 times, or 81 per cent., that 3 times. In the Sartor Resartus (1831) of Thomas Carlyle, the relatives _ which and that occur in all 393 times. Wich is used in restrict- ive clauses 259 times, or in 66 per cent. of all cases, that 134 times, or in 34 per cent. of all cases. The relative in co-ordinate sentences is which, occurring 34 times, and is universally employed. Which is the relative employed with nouns, as in the King James version of the Bible, in about 243 instances, or in go per cent. of all cases. What is used as a relative 93 times, that which 4 times. In Emerson’s Essays, second series (1844), the relatives which and ¢hat occur 402 times; of these, w/ich is used in restrictive clauses 344 times, or in about 86 per cent.; ¢ha# is used in restrict- ive clauses 58 times, or in 14 per cent. of all cases. Wich is used in co-ordinate sentences 27 times, or in all cases, /Aa¢ not at all. What is used 55 times, that which 21 times. Which nearly always relates to nouns, namely, in 330 out of 344 instances of its use. Matthew Arnold, in his Zssays on Criticism (1865), shows a uniform preference for wich in both restrictive and co-ordinate clauses, greater variety and a more flexible adoption of one or the other relative. In four essays, namely, those on ‘‘ Heinrich Heine,’’ ‘‘A Guide to English Literature,’ ‘‘A French Critic on Goethe ’’ and ‘‘ George Sand,’’ in 201 cases of the uses of the rel- atives which and ¢hat, these pronouns are used in restrictive clauses 188 times. Which is used in 186 instances, or in about 288 HEWETT—PRONOUNS IN ENGLISH LITERATURE. [April9, 99 per cent. of all,cases, ‘hat 2, or in 1 per cent.; which is used in co-ordinate clauses 12 times, ‘haf once. What is used as a rela- tive 68 times in the same essays, that which 4 times. The use of what as a relative shows a steady and remarkable growth in frequency in later writers. Its use by Matthew Arnold in the above selections occurs 68 times, or in 25 per cent. of all cases of the use of a relative pronoun. Which is the common: relative in co-ordinate clauses, being used in about 92 per cent. of all the cases. The striking frequency of which in modern literature ‘is shown in» the writings of Mrs. Humphry Ward. The conclusions reached in our examination of the works of Macaulay and De Quincy are maintained, though not in as extreme a degree. Thus in Rodert Elsmere (1888), Book 1, in about one-fourth of the volume, the relatives shat and which occur 400 times; of these, zwAich occurs 350 times, or about 87% per cent., ‘hat 50 times, or 12% per cent. Of relatives referring to noun antecedents which is used 341 times, to pronoun antecedents g times; ¢a? is used referring to a noun antecedent 41 times or 82 per cent., to a pronoun anteced-. ent 9 times or 18 per cent. Mrs. Ward’s use of these relatives is apparently confined to restrictive clauses. Proverbs which have existed in the popular benigsihze for many centuries have preserved an archaic type of expression and are per- manent representatives of primitive usage. Similarly children’s rhymes, such as ‘‘ The house ¢ha/ Jack built,’’ which goes back to a medieval Hebrew version in a hymn. In ‘This is the house that Jack built,’’ ‘* This is the malt /Aa¢ lay in the house ¢ha¢ Jack built,’’ we have the early use of the relative /ia¢ in restrictive clauses ; so also, in such proverbs as ‘* Handsome is /ia¢ handsome does,’’ quoted from Goldsmith in the Vicar of Wakefield, chapter first; ‘‘ He ¢haf will not when he may, when he will he shall have nay’’; ‘‘ There is none so blind as they ¢a¢ won’t see’’; «¢’'Tis an ill dog ¢haz is not worth whistling for.’’ We thus see that the dominant relative f2 of early English times was displaced by /ia¢ in the thirteenth century, that whaz also ap- peared at that time in isolated cases in its relative use, while who and whose occur but seldom and then usually in direct and indirect questions. At the close of the fourteenth century, /ia¢ was used in Wiclif’s translations of the Gospels in 76 per cent. of all cases of the use of the relative, which in 16 per cent. of such cases. One hundred and fifty years later, in 1526, ha? occurs as a rela- OO EE Oe ee 1904.) PACKARD—OPISTHENOGENESIS. 289 tive in the Tyndale version in only 44 per cent. of all cases, while which has risen from 16 per cent. in Wiclif to 50 per cent. in Tyndale. Which was confined largely to nouns and /ha# to pro- nouns. In the eighteenth century, which declines in use in the classical English of Addison and Steele, while sia? gains slightly in frequency. A more marked change is manifest in the nineteenth century in the English of Macaulay, where which refers to a noun in 99 per cent. of all cases of its use as a relative, constituting a marked feature of his style. In Matthew Arnold, this proportion is preserved ; also, though in a less degree, in the writings of Mrs. Humphry Ward. The present tendency is to subordinate the use of that, perhaps in part due to its use as a declarative conjunction, while who has gained in frequency of use and refers mainly to per- sonal nouns. ; Cornell University, Ithaca, April 9, 1904. OPISTHENOGENESIS, OR THE DEVELOPMENT OF SEGMENTS, MEDIAN TUBERCLES AND MARKINGS 4 ZERGO. BY ALPHEUS S. PACKARD, LL.D. (Received June 15, 1904.) Weismann, in his suggestive Studies in the Theory of Descent (1876), was the first to discuss the origin of the markings of cater- pillars, and to show that in Dez/ephila hippophaés the ring-like spots of the larva ‘first originated on the segment bearing the caudal horn, and were then gradually transferred as secondary spots to the preceding segments ’’ (Vol. 1, p. 277). Afterwards (1881-1890), Eimer’ showed that in the European wall-lizard ‘‘ a series of markings pass in succession over the body from behind forwards, just as one wave follows another, and the anterior ones vanish while new ones appear behind.’’ He speaks . 1« Untersuchungen ueber das Variiren der Mauereidechse,” Archiv /. Naturg., 1881; « Ueber die Zeichnung der Thiere,” Zoo/. Anzeiger, 1882, 1883, 1884; Organic Evolution, London, 1890. PROC. AMER. PHILOS. SOC. XLIII. 177. S. PRINTED SEPT. 29, 1904. 290 PACKARD—OPISTHENOGENESIS. [June 15, of this mode of origin of the markings as the ‘‘ law of wave-like evolution, or law of undulation.”” In confirmation of this process or law he cites the conclusions of Wiirtenberger,’ who had long before (1873) observed that in ammonites all structural changes show themselves first on the last (the outer) whorl of the shell, such a change in the following generations being pushed farther and ~ farther towards the beginning of the spiral, until it prevails in the greater number of the whorls.’’ Cope, in his Primary Factors of Organic Evolution (1896), also shows that in the lizards Cuemidophorus tesselatus and gularis the breaking up of the striped coloration into transverse spots begins first at the sacral and lumbar regions: ‘‘ The confluence of the spots appears there first.’’ We may cite some examples of this law of growth a ¢ergo, or opisthenogenesis, as it might be called, which have fallen under our own observation.’ In Dasylophia anguina, as shown by the figures in Plate XXI of my monograph of the bombycine Moths, Pt. 1, it will be observed that in stages III, IV and the last stage the dark longitudinal lines become on the eighth to tenth abdominal segments broken up into separate isolated dark spots. In the larva, before the second molt, there are no spots on the ninth and tenth segments. In stage III, however, z.¢., after the second change of skin, as stated in my monograph (p. 175), four black spots now appear on the front part of the suranal plate. In the last stage, the reddish spots on the eighth abdominal segment which are detached from the lateral lines of stages I and II, now become specialized into the two black comma-like spots, with a linear spot above and beneath ; the two, sometimes divided into four, black spots arise on the suranal plate. It thus appears that in the ontogeny of this species the process of breaking up or origin of the spots from the longitudinal lines takes place on the last three segments of the body. In Symmerista albifrons the same phenomenon occurs, In stage I, as stated in my monograph (p. 180), on each side of the ninth segment is a large black comma-shaped spot, the point directed ‘ forward and downward, while behind there is a median black dot. 1A New Contribution to the Zoological Proof of the Darwinian Theory, Ausland, 1873, Nos. 1, 2, and Studies on the History of the Descent of the Ammonites, Leipzig, 1880 (in German). * Proc, Amer, Asso, Advancement Science, Boston Meeting, 1898, pp. 368-9. — 1904. PACKARD—OPISTHENOGENESIS. 291 After the first molt there arises behind the dorsal hump two, instead of one, median black spots, and two black spots are added on the side of the body near the base of the anal legs, z.e., two each on the ninth and last segments. After the second casting of the skin, the marking of the three last abdominal segments becomes specialized ; what on the body in front are parallel black and red lines being in this region now represented by separate spots. Thus as regards the marking, the anterior part of the body remains ornamented with the primitive parallel lines; while the process becomes on the three hinder seg- ments accelerated or specialized. It thus appears that the more advanced or ontogenetically later style of ornamentation originates at the end of the body. A parallel process takes place with the formation of the caudal horn or hump. Thus in Symmerista, Dasylophia and other horned Notodontide and members of other groups, the eighth abdominal segment is the theatre of the process of fusion of the two dorsal tubercles of the first larval stage into a single tubercle or horn ; so that this segment appears to be the theatre of a process of specialization which does not take place on any other segments of the body. When in other genera it does take place and there is a special- ized single tubercle on the first abdominal segment, as in Noto- donta, Nerice and more especially in Hyparpax and Schizura, the process of fusion of two tubercles into a single specialized one, as on abdominal segments 1, proceeds from behind forward, as it were in waves of translation of the specialized growth-force from behind forwards. This may clearly be seen in the figures on Plate XXIV, showing the development of the single hump in Ayparpax aurora. In Fig. 1, the dorsal tubercles 7 in stage I are all separated; in Fig. 2, those on the eighth abdominal segment have all begun to unite at their bases before they have on the first abdominal segment; they seem to be a little behind at first, though later on the hump on the first segment becomes higher and larger than the caudal horn. If there were any doubt as to the relative period when the tubercles become fused in Hyparpax, in Schizura leptinoides (PI. XXVI) it is very clearly shown by Fig. 1 that the fusion of the two tubercles forming the caudal hump as we will call it, ¢., that on the eighth abdominal segment, has taken place before any signs 292 PACKARD—OPISTHENOGENESIS. [June 15, of such fusion have appeared in the pair on any of the segments in front. When the ontogeny of Verice didentata is worked out, it will be a matter of much interest to observe whether the dorsal humps are formed from behind forward, or whether they appear simultaneously, and thus form an apparent exception to the law of transfer of growth-force from behind forwards. In this connection it might be observed that in the larva of Schizura unicornis,in which there is the very unusual occurrence of a pair of short thick spines on the vertex of the head (Pl. XXVIII, Fig. 2, 2a, 2b), these spines do not appear in stage I and not until after the first molt. These spines persist through stages II and III, but after this disappear, not being present in the two last stages. Thus the growth-force resulting in the development of the armature of stage I does not reach the head until after the first molt, and then does not persist throughout larval life. In the ontogeny of the Notodontian family, as well as that of Ceratocampidz and Saturniid, the process of fusion of the two dorsal tubercles always first begins on the eighth abdominal segment. Opisthenogenesis, as regards ‘the markings, appears to be of a piece, or somehow connected, with the opisthenogenetic origin in post embryonic development of new segments. In the cestodes and in annelid worms, multiplication of segments occurs between the head-region and the extreme end of the body. Thus in Poly- gordius, as stated by Balfour (4 Treatise on Comparative Embry- ology, 1880, I, pp. 271, 272), the conversion of the larva into the adult takes place ‘‘ by the intercalation of a segmented region between a large mouth-bearing portion of the primitive body and a small anus-bearing portion.”’ This region in the larval or early stages of worms and more primitive arthropods is the ‘‘ budding zone’’ of embryologists. While at the outset, in the beginning of embryonic life, the head - region is the first to be formed and the trunk-segments arise later, as in the trochosphere of worms and the protaspis of trilobites and of merostomes, a third portion, arising from the budding zone or seat of rapid cell-formation, appears to be a secondary or inherited region, due to the post-embryonic acquisition of new characters (certain trunk-segments and their appendages) in many segmented or polymerous animals, #.¢., those which have passed beyond the trochozoén stage or type. Oa eee. Goan 1904.] PACKARD—OPISTHENOGENESIS. 293 Prof. E. B. Wilson! has clearly stated the nature, now so well known, of the growth-processes involved in the interpolation at the growing point or budding-zone of new segments. In Polygordius, after the trochosphere has been formed and when it is about to enter on the adult stages, the segments are formed successively, those in front being the oldest, ‘‘ while new segments are contin- ually in process of formation, one after another, at the growing point.’’ This, he says, is ‘‘a typical case of apical or unipolar growth.’’ It is what we would call opisthenogenetic growth. Professor Whitman’ has shown that in the leech the internal tissues (mesoblast) of the budding zone are arranged in two widely separated lateral bands which, to quote Wilson’s exposition, ‘‘ as the trunk grows older, widen out and grow together along the me- dian line, ultimately giving rise to muscles, blood-vessels, excretory organs, reproductive organs, etc.’’ Now if this is the case with the more important tissues, why in caterpillars as well as in lizards may not this opisthenogenetic mode of growth also involve the arrange- ment and distribution of the pigment-masses of the integument ? Without entering into the mode of development of the germ- bands, which are behind completely separated, gradually becoming united in front, resulting in their union or concrescence, we would make the suggestion that this phenomenon may be the initial cause or at least in some way connected with the breaking up of the lon- gitudinal stripes of the body, and their transformation into spots at or near the budding zone of their polymerous or polypodous (Peri- patus-like) ancestors. In the trilobites, Limulus and Diplopods, the new segments after embryonic life are interpolated between the penultimate and anal or last segment of the body, and it is from this region in certain Lepi- dopterous larve that the transformation of longitudinal stripes into spots takes place. The question next arises whether there is any connection between the opisthenogenetic origin of the markings of lizards and that of caterpillars. The fact, now well established by embryologists, that the phenomena of concrescence occurs not only 1 Some Problems of Annelid Morphology. Biological lectures delivered at the Marine Biological Laboratory at Woods Holl, 1891, p. 61. See also A. D. Mead, “ The Early Development of Marine Annelids,” Yournal of Morphology, XIII, May, 1897, pp. 227-326. 2« The Embryology of Clepsine,” Yourn. Micr. Se., XVIII, 1878; Fournal of Morphology, Boston, 1887, I am indebted to Prof. A. D. Mead for calling my attention to the concrescence process in this connection. 294 BROOKS—ORTHIC CURVES. [May 20, in fishes but in Amphibia and reptiles, would suggest that the cause of the transformation of longitudinal stripes into spots on the lum- bar and sacral regions of lizards is the result of the same specializing growth-force. It may perhaps be regarded as a surviving remnant of the segment-forming force, which has affected the pigment bands in a manner identical in the vertebrates and insects. This trans- formation of stripes into spots, and the fusion of two dorsal tubercles into a median one, may be, then, the sign of some latent or surviving amount of force concerned in the origin and forma- tion of segments, which crops cut in the larval stages of insects and in young lizards, resulting in this opisthenogenetic mode of origin of spots from bands. ORTHIC CURVES; OR, ALGEBRAIC CURVES WHICH SATISFY LAPLACE’S EQUATION IN TWO DIMENSIONS. BY CHARLES EDWARD BROOKS, A.B. (Read May 20, 1904.) I propose a study of the metrical properties of algebraic plane curves which are apolar, or, as it is sometimes called, harmonic, with the absolute conic at infinity. If we disregard the right line, the simplest orthic curve is the equilateral (conic) hyperbola, and the name equilateral hyperbola is sometimes extended to orthic curves of higher order. Doctor Holzmiiller,! who devotes a section to curves of this kind, calls them hyperbolas; and M. Lucas’ calls them “stelloides.’’ M. Paul Serret, in a series of three papers in Comptes Rendus,* uses the word ‘* équilatére’’ for a curve with | Einfiihrung in die Theorie der Isogonalen Verwandschaften und der Conformen Abbildungen, Gustav Holzmiiller, Leipzig, 1882, p. 202... . *« Géométrie des Polynomes,” Felix Lucas, Journal de l’Ecole Polytech- nique, 1879, t. XXVIII, ; * Comptes Rendus, 1895, t. 121. Sur les hyperboles équilatéres d’ordre quelconque, p, 340. Sur les faisceaux regulieres et les équilatéres d’ordre n. p. 372. Sur les équilatéres comprises dans les equations Om 5, -8/, 7% H,, O = Y,2-1/, 7,9 = Hat AH. p- 438. SOE rr, 1904.] BROOKS—ORTHIC CURVES. 295 asymptotes concurrent and parallel to the sides of a regular polygon. It seems advisable to follow M. Serret’s usage, and to denote such a curve by the name equilateral, using another term to express apolarity with the absolute. For this purpose I have adopted the word orthic. If we use Cartesian codrdinates, a curve : KHXY) =O, is apolar with the absolute conic, 3 + 4°? =0, if aU aU s gu + oy =? In other words, an orthic curve is one which satisfies Laplace’s equation in two dimensions. Part ONE—THE Ortuic Cusic CURVE. I. Zhe Condition that a Curve be Orthic. In the analysis which may be required, I shall employ conjugate coérdinates, x, x, which may be defined as follows: If X and Y are rectangular Cartesian codrdinates of any point, the conjugate coérdinates of that point are x= X+iV,x=—X—1Y, when the origin is retained, and the axis of X is chosen as the axis of reals, or base line. It is sometimes convenient to think of x as the vector from the origin to the point, and of x as the reflection of this vector in the base line. If x, x is a real point of the plane, not on the base line, «— x0, x and x are conjugate complex numbers. Since if one of its codrdinates is known the other is immediately obtainable, we shall, as a rule, name a point by giving only one of its codrdinates. It is convenient to reserve the letters #and +t for points on the unit circle, ee oe 3S Now, Laplace’s equation, aU BU ant oY = oO 296 BROOKS—ORTHIC CURVES. . [May 20, when. applied to a function of « and x, becomes U(x) ss axax It follows that : The necessary and sufficient condition that a curve be orthic is that its equation in conjugate codrdinates contain no product-term. Il. Xinematical Definition of the Orthic Curve. Let us now proceed to the study of the orthic curve of the third order. I shall obtain the equation of an orthic cubic in a way which will suggest immediately a method for the construction of points on the curve. The path of a point which moves in such a way that tt preserves a constant orientation from three fixed points ts an orthic cubic curve. If x is the moving point, and the three fixed points are a, f, 7, then the sum of the amplitudes of the strokes which connect x with a, 8, 7, must remain constant. That is, we must have (# — a) (« — 8) (x—7) =r, If the curve is to be real, the conjugate relation, (%#—a) (eB) (®#—7) = er", must hold simultaneously. The equation of the curve is obtained by eliminating the para- meter p between these. It is xe —(a-+ 8+ 7) x*-+ (a8 + Ay + ya) «x —afy =n {xt— (a+ B+ 7) + (a8 + By +70) © — 287}. This is the most general equation of the third degree which we can have without introducing the product. As a consequence it represents a perfectly general orthic cubic. If we transform to x=}(a++y7), the centroid of afy, as a new origin, and so choose the base line that +," is real, the equation takes the form aS ae arteritis 1 ne o- 1904.) BROOKS—ORTHIC CURVES. 297 at + aye + a+ ax + x*=0. The equation of any orthic cubic can be brought to this form. The three points, a, # and 7, are on the curve, and form what it is convenient to call a triad of the curve. Ill., Zhe Orthic Curve is an Equilateral Curve. Consider the orthic cubic, a* — $,x* + $9" — $4 =7) (x? Ml 58 a Sy —4$,); where the s’s are the elementary symmetrical functions of a, 4 The approximation at infinity, (x — $5)" — (x =a 45,)* = 0; makes both the square and the cube terms vanish, and therefore represents the asymptotes. ‘The factors of this are: x — $s, — #7," (x — $5.) = 0, x — 4s, —w.¥r} (x — ts,) =o, x— 45, — 0. Yr) (x — 4s,) =o. where w* = 1. These three lines meet at the point x=4%(¢+8+7) which we may call the centre of the curve. We notice that: The centre of the orthic cubic is the centroid of the triad. The clinants of the asymptotes are 7,8, wr,3, w*r,3. They differ only by the constant factor w. Now we know that multiplying the clinant of a line by is equivalent to turning the line through an ar 27 angle om A rotation ss about the centre sends each asymptote into another. It follows that the asymptotes of an orthic cubic are concurrent and parallel to the sides of a regular triangle. M. Serret’ calls such a figure of equally inclined lines which meet in a point a regular pencil, and a curve with asymptotes forming a regular pencil he calls an ‘‘ éguclatére.”’ 1 Comptes Rendus, Sur les hyperboles équilatéres d’ordre quelconque. 1895, t. 121, p. 340. 298 BROOKS—ORTHIC CURVES. [May 20, Now any cubic curve, the asymptotes of which form a regular pencil, can be brought to the form : w+ aye + a, + Ox +x =0, in which we recognize it as orthic. It follows that: The orthic cubic and the equilateral of order three are identical. The relation (@— a) (x —8) (e—y) =p = may be regarded as mapping a line through the origin in the z plane, re \ s—t,z=0 into the orthic cubic. We are thus able to identify the latter with the curves discussed by Holzmiiller’ and by Lucas.’ IV. Construction of Points of an Orthic Cubic. A figure of the orthic cubic may be obtained without great diffi- culty by constructing points of the curve. In order to show how this may be done, it is necessary to prove the following lemma: Elements of the pencil of equilateral (orthic) hyperbolas, of which the stroke fy is a diameter, intersect corresponding elements of the pencil of lines through «on an orthic cubic of which aBy ts a triad. For the line through a, (w— a) =pr, and the equilateral hyperbola on fy as a diameter, (x —f) (w—y)= pr", intersect on the orthic cubic (x— a) (wv —8) (@—7) =r 'Holzmilller, Conformen Abbildungen, p, 205. *Lucas, Géométrie des Polynomes, Journal de I’ Ecole Polytechnique, t. XXVIII, p. 23. 1904.] BROOKS—ORTHIC CURVES. 299 if dd! =n. If the two pencils are given, it is only necessary to pair off lines and curves according to the relation v=, and to mark intersections. These will be points of the curve. _A very simple instrument for drawing the equilateral hyperbolas required in the construction is made in the following way: ‘Two toothed wheels of equal diameters are attached beneath the drawing Figure 1. A unipartite orthic cubic which has three real inflections, one of which is at infinity. board in such a way that their teeth engage. The axles are perpen- dicular to the board and come through it at Pandy. The axles, which turn with the wheels, carry long hands or pointers which sweep over the board. On account of the cogs, the wheels can turn only through equal and opposite angles. As a consequence, .v, the 300 BROOKS—ORTHIC CURVES. [May 20, intersection of the hands, has a constant orientation from and 7, and in fact generates the orthic curve of the second order given by (x — 8) (x —y) =r, which is the hyperbola required. V. Mechanical Generation of an Orthic Cubic. A mechanism which will actually draw an orthic cubic is very much to be desired. One might be made in some such way as the following: Suppose three hands like those described above (IV) to be pivoted at a, Sandy. Let them be held together in such a way that, while each is free to move along the others, they must always meet in a point, which is to be the tracing point. Each hand is to receive its motion from a cord wound about a bobbin on its axle. The bobbins are to be equal in diameter. The cords pass through conveniently placed pulleys, and are kept tight and vertical by small equal weights at their ends. Consider, to fix ideas, those three weights which by their descent give the hands positive rotation. If, now, the tracing point be moved along an orthic cubic which has a, #, 7 for a triad, the total turning of the bobbins will be zero, and as a consequence the total descent of the weights will be zero. Conversely, if we can move these vertically and in such a way that the total descent will be zero, the tracing point can move only along an orthic cubic. This result will be obtained if the centre of gravity of the three weights can be kept fixed. It will not do, however, to connect the three weights by a rigid triangle pivoted at its centre of gravity, for then they will not move ver- tically. But since a parallel projection does not alter the centroid of a set of points, the desired result will be attained if the weights are constrained to vertical motion by guides of some kind, and are kept in a plane which always passes through the centre of gravity of one position of the weights. VI. Zhe Orthic Cubic through Six Points of a Citrele. Consider the general orthic cubic given by x? — a,x’ + ax —a, +a. — a + a,x"'= oO. It cuts the unit circle, 1904.] BROOKS—ORTHIG CURVES. 801 in six points, the roots of x — ayx® + a,x* — a,x* + a,x* — a,x + a,= 0. If we want the cubic to meet the circle in six given points, say 7, T;,. . . T., then this equation must be identical with x — sya + 5,04 — 54x° + 54x" — 55x + 56 = 0, in which the s’s stand for the elementary symmetrical combinations of the six 7’s. This requires Ay = Sy, A = Sq, Ap = Sy As = Sq, Ay = S5, As = Se The coefficients of the cubic equation are then precisely determined, with the result that: But one orthic cubic can be constructed through any six points of a circle. It remains for us to show that one such curve can always be drawn: that is, that the equation a2 — $22 + sox — 55 + 5x — 5,x° + sex? = 0 always represents a real curve. If we so choose the base line that Sg == 1 then we have 5; Sg Sgn Sos and the equation takes the form x — 5,27 + 59x — 5,4 54x —5,0°-+ 2? = 0, which is, obviously, self-conjugate, and is therefore satisfied by the coérdinates of real points. As a result: An orthic cubic can always be drawn through six points of a circle. Lt ts then determined uniquely. VII. Zhe Lntersections of an Orthic Cubic with a Circle. When the orthic cubic is referred to the six points in which it cuts the unit circle, the equations of the asymptotes take the form s—45,=(— 5s) Ayy' (x — 4 5556-*). SSO) Fas 302 BROOKS—ORTHIC CURVES. [May 20, These three lines meet at x=45, the centre. This point, the origin, and the point which is the centroid of the six points on the circle lie on a line ; and the latter point is midway between the other two. This leads to the interest- ing fact that : The centroid of the six points in which any circle meets an orthic cubic bisects the stroke trom the centre of the curve to the centre of that circle. VIII. Zriads of the Curve. We spoke of the three points a, 8, 7, which have the same orienta- tion from every point of the curve, as a triad of the curve. Let us see how many such triads there =e) and how they are arranged. The relation (x — a) (*@—8) (x—y) = may be regarded as establishing a correspondence between points x in one plane and points z in another plane, in such a way that if z describe a line € through the origin, the point x generates an orthic cubic on afy asa triad. To every position of z on the director line € there correspond three points in the x-plane. I shall show that each such set of three points is a triad. Write F (x) = (x — a) (w— A) (x—7). Then, if x,, x,, xs, are the three points which correspond to sz, F (x) —2=(x—«x,) (xn — x,) (x — x). And also F (x) — 2! = (« — x) (x — x’) (x — x’). Now this relation is satisfied by x,, or x,, or 5. F (x) —7 = (%,— x’) (4 — ay) (4, — x) = — 8. Since z —2z’ is a point of the director line, it follows that the three points «,', x,', x, which correspond to any point ¢’ of the director line, have the same orientation from every point of the curve. We conclue that; To every point of the director line corresponds a triad; all the 1904.] BROOKS—ORTHIC CURVES. 308 points of the curve have the same orientation from any triad, and all the triads of the curve have the same orientation from any point of the curve. IX. Zhe System of Confocal Ellipses Connected with the Triads. We seek the points of a triad which correspond to a given point z.. The map equation can be brought to the form x? — 3x == 28 by choosing the centre of the curve as a new origin and making a suitable choice of the unit stroke. We see at once that the sum of the x’s for a given z is zero. In other words: Zhe centroid of any triad ts the centre of the cubic. Making use of the method known as Cardan’s solution, put x= pt+ 2, ~where » is real. Then x*— 3x = 22 becomes we + + nity + gute” — 3 (ut + v) = 22 And we have as two relations between v and y/, 22 = p+, and (ut+ v) (utv —1)=0. When 2 is zero, the values of x are + 1/ 3 and 0; and when z is not zero, we must have I v= be This leads to the expression of x and z in terms of u# as follows: —— ft I x= ph-+ iW I 2g = Wl + Tape Now if we assign any value to #, and let ¢ run around the unit circle, x describes an ellipse with foci at «= - 2 and x= —2. 1 Harkness and Morley, A Zreatise on the Theory of Functions, p. 39. 304 BROOKS—ORTHIC CURVES. [May 20, But at the same time, z also describes an ellipse with its foci at z=-+1andz——1. These two ellipses are related in such a way that a point z on one of them is correlated by the equation x — 3x = 22 with three points on the other. Now the foci of both these ellipses are independent of the particular value of » selected ; it follows that if we assign successive values to », we shall obtain in each plane a system of confocal ellipses of such a sort that the equation x*§ — 3x = 22 establishes a one to one correspondence between them. In each plane the origin is the centre of all the ellipses. Applying this scheme to the case in hand, we see that a triad must be inscribed in one of the ellipses in the x-plane. But the centroid of the triad is the centre of the ellipse ; so the ellipse must be the circumscribed ellipse of least area of that triad. We may say, then, that: The triads of the orthtc cubic are cut out on the curve by a particu- lar system of confocal ellipses, and each ellipse is the circumscribed ellipse of least area of the triad on it. X. Zhe Riemann Surface for an Orthic Cubic. If we examine the equation x*— 3x = 2 for equal roots, we find that the double points of the «-plane are at x==-++1and at x=—r1. These values of x correspond to the branch points in the z-plane, s = +-1 and s=—1. Let us for a moment replace the z-plane by a three-sheeted Riemann surface. All three sheets must hang together at infinity, and two sheets at each of the branch points. Let the first and second sheets be connected by a bridge along the base line from -+- 1 to infinity, and the second and third sheets be similarly con- nected by a bridge along the real axis from — 1 to infinity. Select on this surface any large ellipse with foci at the branch points, and any line as a director line. Now consider the contour obtained by starting from a point of this inside the ellipse, going thence along the line to meet the ellipse, along an arc of the ellipse to meet the line, and then along the line to the point of departure. 1904.] BROOKS—ORTHIC CURVES. 305 We can choose this path in such a way that one of the following three cases must arise : (1) Zhe contour passes through a branch point. (2) The contour surrounds two branch points. (3) The contour surrounds no branch point. In case (1) we know that the cubic must have a node. In the second case, by going three times around we can pass continuously through every sheet of the Riemann surface and therefore through every value of x. Or, thinking again of the x-plane, we have a unicursal boundary. Now it happens that the ellipse we choose maps into one and not three ellipses on the x-plane. If we imag- ine this to expand indefinitely we shall have to consider the bound- ary as our orthic cubic. It follows at once that: The orthic cubic which corresponds to a line which does not pass between the branch peints ts unipartite. If the contour includes one branch point, and therefore crosses one bridge of the Riemann surface, we must go along two uncon- nected curves to reach all the values of x. When these two curves are spread on the x-plane they lead at once to the conclusion that: The orthic cubic which corresponds to a line which passes between the branch points ts a bipartite curve. XI. Triads in Special Cases. Let us turn our attention again to the two planes connected by the relation x* — 3x = 22 We notice that while the ellipses in the z-plane have their foci at the branch points, the foci of the corresponding system of ellipses are not the double points of the x-plane, but are the points x==-+ 2 and x=—2z, each of which, with one of the double points counted twice, forms a triad. As a rule there are two triads of the curve on each ellipse, corre- sponding to the two points in which the director line cuts an ellipse of the system in the z-plane. But unless the line go between the branch points it will be tangent to one ellipse, consequently two triads will coincide, and the cubic will be tangent at three places to one of the ellipses of the system. No part of the cubic can be inside of that ellipse. » PROC. AMER. PHILOS. SOC. XLII. 177. T. PRINTED SEPT. 29, 1904. 306 BROOKS—ORTHIC CURVES. [May 20, When z is 1, the two ellipses degenerate into two segments, ' g=t+ or 2,—2,° ae f 4 e=* OF ty Ke If the line pass between the branch points, and so cut the seg- ment 1, — 1, two triads again coincide, but in this case the three points lie on a line, and we do not have the triply tangent ellipse. When the line & cuts the axis of imaginaries, 7 z+2z=0, we have ml o= p-e, and ri f= p's, It follows that am ¢—= 7, and so of is the reflection of ¢ in the axis of imaginaries and *f is a pure imaginary. Then, since we know that I ee me OAL rn Sage $= TI, 2, 3; we see that x, is the reflection of x, in the line « + = o, and that x, is on that line. It follows that the triangle x,x,x, is isosceles and that its base x,x, is parallel to the real axis. There is again an isosceles triangle when /* is real. This triangle has its vertex on the axis of reals and its base perpendicular to that axis. From the discriminant of the quadratic in y'¢*, s— 4, we see that /* is real when z> +1. In other words, if the director line € cut the axis of reals, but not between the branch points, we have such an isosceles triangle. From the above considerations, we see that if the director line is either of the axes x-+x=0,*x—x=0, then one branch of the orthic cubic must bee right line; the re- 1904.] BROOKS—ORTHIC CURVES. 807 maining portion of the curve must then be an ordinary hyperbola, and the inclination of its asymptotes must be either 7 or #*. The first value refers to the case when the director line is the axis of imaginaries ; and the last, to the case when it is the axis of reals. XII. Zhe Intersections of the Circumscribed Circle of a Triad with the Cubic. Suppose we put a circle through the points of a triad, and ask, Where are the remaining three points in which it cuts the cubic? For convenience, let three points of the unit circle be taken as a triad. The cubic is then (@—4) (@—A#) @ A) =n E—AD F—4) F—4. On eliminating x from this and the equation of the circle we obtain tT? (4—+*) (4—+) (4—) (x— 4) («—4) («—4) = Oe 2 or —T,2 gat, Alat as the equation of the three points sought. The roots of this, By aS Ay Bg TS 8, Ky we, are the codrdinates of the vertices of an equilateral triangle. As there is no restriction in taking the triad on the unit circle, we have the following theorem : If a circle cut an orthic cubic in a triad, then the two curves have three other intersections, which form an equilateral triangle. XIII. Zhe Pencil of Orthic Cubics Which Have a Triad in Common. We have seen that the relation (w— a) (*—8) (x—y) =2 maps a line through the origin into an orthic cubic of which ay is a triad. It must then map all the lines through the origin into a single infinity of orthic curves* which have the common triad afy. Felix Lucas, ¥ournal de? Ecole Polytechnique, t. XVIII, p. 21. 308 BROOKS—ORTHIC CURVES. [May 20, If we regard t as a parameter, we may say that (* —a) («—f) (w®—7)=t (%—2) (« —A) (x—7) is the equation of the pencil of orthic cubics which have the triad afy. It will be convenient to give a pencil of this sort some name; let us refer to it as a central pencil, noting for our justification that the centroid of the triad is the centre of every curve of the pencil. If there were any real point other than a, f, ory, on two curves of this pencil, it would map into a real point of the z-plane, not the origin, which would be on two of the lines through the origin. As this is manifestly impossible, it follows that: Two orthic cubics which have a triad in common have no other real intersection. Now we know that two cubic curves intersect in nine points, and that if the curves given by the equation (a — a) («—f) (w—y) =t (w@—a) (x —8B) (x —7) really constitute a pencil, there must be six imaginary points whose codrdinates satisfy this equation whatever the value of rt. Let us form the following table of codrdinates. The real intersections are 4, == 4, x =a, y= B, m= 8, xX,=?); Xs = y: It is evident that each of the following points : i= a, x, =B, X,= 4, %=y, Xx =f, a= 4, ty == By ty = r t=) m=, y=" Es =f, satisfies the equation, independently of r. These points, the six imaginary intersections of the pencil, are the antipoints' obtained 'Cayley, Collected Mathematical Papers, Volume VI, p. 499. 1904.] BROOKS—ORTHIC CURVES. 309 by selecting pairs in all possible ways from a, f, 7 The figure of nine points in which two orthic cubics intersect may be regarded as an extension of the orthocentric four-point determined by two equilateral hyperbolas. It is convenient to extend the term orthocentric to such a figure. Resuming the results obtained above, we have: When three of the points of an srebacendiee nine-point are a triad of the orthic curves through the nine points, the remaining six points are imaginary, and are the antipoints of the three real points. The centroid of the nine points ts the centre of every orthic cubic through them. It is convenient to speak of a set of orthocentric points deter- mined by a central pencil as a central set. Since any three points determine a pencil of orthic cubics of which they are a triad, any three points, with all their antipoints, form a central orthocentric nine-point. XIV. The Foct. We shall now attack the problem of finding the foci of the orthic cubic. Let us begin with a few words as to the way in which the foci of a curve appear in analysis with conjugate codrdinates. The focus of a curve is the intersection of a tangent from one ciroular point with a tangent from the other circular point. In other words, if the circular rays from a point are tangent to a curve, that point is a focus of the curve. Now the equation of the circular rays from a point a, a, is' ee (x—a) = 0. Therefore, one of the lines is , x—a=0, and the other is %—a==0. Suppose the equation of the curve is F(xx)=0. ' Now if the circular ray . x— ao 310 BROOKS—ORTHIC CURVES. [May 20, is tangent to the curve, then Flax) == 0) the eliminant of x between these two, wiil have equal roots. But since the equation of a real curve must be self-conjugate, if this has two coincident roots, then _— Fax) = 0 must also have, and thé point a, a, is a focus. It follows that to find the foci of a curve, we have merely to find those values of x which make two values of x coincide. They are the vectors of the foci. Let us apply this method to the orthic cubic. The equation may be taken in the form e@— 32 = 23 =a,+ Ay, where A is a real parameter and the director line is a, + Aa, = 232, a + da, = 23, These relations imply the conjugate expression a — 30 = 22 =a, + da, Two values of x become equal when Dzz = 0, #.e.,"when #—1=0, or w= +1, These values of x occur when a, + da, = + 2, or 4) +2 a A= Either of these values of A when substituted in a — 34 == a, -+- Ad, gives three points which are foci of the cubic. eh Ce sie ee Meee 1904.] BROOKS—ORTHIC CURVES. 811 There are, in general, six real foci, which fall into two sets of three. Lach set of three corresponds toa single point of the z-plane, and ts, therefore, a maximum inscribed triangle of one of the ellipses described above. * XV. Zhe Foci and the Branch Points. If we eliminate the parameter between 22a, +g and 23 =a, + das, we get the equation of the line é, az ee, a3 = a; a yA. Now suppose, for a moment, that this line does not contain either of the branch points z=+1. Then, if we put z—-+1 in the equation of the line and solve for z, we get a value which is not the con- jugate of 2, but is the vector of the reflection of the point s—=+1 in the line considered. The three points in the x-plane got by putting in the equation x — 3x — 22 are the points mapped in the z-plane by the reflection of z+ 1 in the line €&. It follows that: The real foci of the orthic cubic which corresponds to a given line are the six points which correspond to the reflections in that line of the branch points. If the director line pass through one of the branch points (¢.¢., if = is real), two foci coincide to form the node, and the 1 remaining one of the set is on the curve. One who looks at the matter from the point of view of the Riemann surface might be surprised that a branch point is to be reflected in the line in each sheet of the surface and not in the two sheets alone which it con- nects. A moment’s consideration will show that whether or not S{2 BROOKS—ORTHIC CURVES. [May 20, two x’s coincide depends on / alone, and that either of three values of x give 2a particular value. It is clear that the reflection must be in every sheet of the surface. In general, the orthic cubic is of class six. Since it cuts the line at infinity in three points apolar with the circular points, it cannot contain one of the circular points unless it is as a point of inflection. There should, therefore, be six tangents from each of.the circular points and, consequently, thirty-six foci. The thirty foci still to be accounted for are the antipoints' of the six real foci, paired in all possible ways. When the cubic has a node it is of class four, and has but four real foci. The node takes the place of the two foci which coincide there. The circular rays x—a,=0 and X— =O meet at a,, a. So the thirty-six foci of an orthic cubic may be represented by the scheme of coérdinates : Gi, Ay where 7 and / run from one to six. It follows that the centroid of the whole thirty-six is the centroid of the six real points; that is, the centre of the cubic. Consider any selection of three foci. All their antipoints are foci, and the nine points together make up a central orthocentric set. XVI. Zhe Foci of the Orthic Cubics which Have a Triad in Common Lie on Two Cassinoids. The foci of all the orthic cubics which have a commormtriad apy le on two cassinoids which have their foci at a, 8, and y, and are ortho- gonal to the orthic curves. We know that these cubics correspond to all the lines through a point, and that their foci correspond to the reflections of the branch points in those lines. Now the reflections of a fixed point in all the lines through a second point lie on a circle which goes through the first point, and which has its centre at the second point. Accord- 'Salmon, Higher Plane Curves, third edition, p. 122, 1904.] BROOKS— ORTHIC CURVES. 318 ingly, the foci of the cubics will lie on the curves which are the maps in the x-plane of two concentric circles in the z-plane. The centre of these circles maps into the triad common to all the cubics, and the circles themselves map into two cassinoids of the sixth order, about the triad, as M. Lucas'has shown. Each of the circles goes through one of the branch points, and, therefore, each of the cassinoids must have a node. If the point which corresponds to the triad afy is equidistant from the branch points, the two circles and also the two cassinoids coincide. In this case the cassinoid has two double points. The lines which correspond to the cubics are all perpendicular to the circles which correspond to the cassinoids, and so, by the prin- ciple of orthogonality, the ovals are orthogonal trajectories of the cubics of the pencil. XVII. Zhe Position of the Orthic Cubic in Projective Geometry. I shall close this study of the metrical properties of the orthic curve of the third order by showing that from the point of view of projective geometry the orthic cubic is really a general cubic. Any proper plane curve of the third order can be projected into an orthic curve. We know that the points of contact of three of the six tangents to a cubic curve from any point of its Hessian lie in a line. Now these three points, considered as a binary cubic, have a Hessian pair. Ifthis pair of{points be projected to the circular points at infinity, the three tangents become equally inclined asymptotes, and continue to meet in a point. The cubic curve is then orthic and the transformation is accomplished. This projection requires two points to go into given points, and can, therefore, always be made. Jn projective geometry the orthic cubic is any proper plane cubic. As an illustration of the way in which information about the orthic cubic applies to cubic curves in general, let us see what the characteristic property that the asymptotes are concurrent and equally inclined means. The circular points /and /are a pair of points apolar with the curve. Their join, the line at infinity, meets the curve in three points such that the tangents at these points meet 1Felix Lucas, Géométrie des ‘ian iowapas Journal de LP Ecole Polytechnique, XXVIII, p. 5. 314 BROOKS—ORTHIC CURVES. [May 20, in a point, C, of the Hessian. Now we know’ that such a line meets the Hessian in the point which corresponds to C. This leads to the theorems that: The line joining two points apolar with a cubic curve meets the cubic in three points, the tangents at which meet in a point of the Hessian, and are apolar with the two points apolar with the curve. The line joining two points apolar with a cubic curve, and a tangent to the cubic at a point of this line, meet the Hessian of the given cubic in corresponding points. A more novel result is the following. We have seen (XIV, p. 28) that the foci of an orthic cubic fall into two sets of three, in such a way that the two sets are triangles of maximum area inscribed in two confocal ellipses. Now if we consider tangents from 7 and J instead of foci, we have the following theorem : Ifaand b are a pair of points apolar with a cubic curve, then the tangents from either of these points, say a. fall into two sets of three in such a way that the line ab has the same polar pair of lines as to each set of three. Part Two—OrtHic CURVES OF ANY ORDER. I. Yntroduction. In the preceding pages we have studied the metrical properties of the orthic cubic in some detail. In the following portion of the work I shall indicate an extension of the more important results obtained in the study of the cubic to orthic curves of any order. The general equation of the ”™ degree between x and x contains Yn(n —1) product terms. If it is to represent an orthic curve the coefficients of these terms must be made zero. In other words, to ' make a curve of the m™ order orthic is equivalent to making it satisfy %4m(m—1) linear conditions. After this has been done there remain 2” degrees of freedom. II. Zhe Orthic Curve is Equilateral. The kinematical definition which we obtained for the orthic cubic may be extended to curves of any order, that is; Salmon, Higher Slane Curves, third edition, articles 70 and 175. 2 On the Algebraic Potential Curves,”’ Dr, Edward Kasner, Budietins of the American Mathematical Society, June, 1901, p. 393. 1904.) BROOKS—ORTHIC CURVES. 815 The path of a point which moves so that its orientation from n fixed points ts constant ts an orthic curve of order n. If a,, a, ... a are the fixed points, the condition on x is expressed by the relations (x—a,) («— a)... (x —a,) =pr, and (%#—%) @— 4G). . « (a) Spr. These lead to the equation of the curve, HP mm $509? 5gee . . pt! (Sa Sea ner —x*) =o, where the s’s are the elementary symmetric combinations of the a’s- This is the general equation of an orthiccurve. If we take x= %s, for a new origin, and make +,’ real, the equation becomes at § — ay... ay tas + =O. The asymptotes are the # equally inclined lines given by the factors of the highest terms, x®* + x°= 0, These lines all pass through the origin; it follows that the centroid of the points a,, . . . a, isthe centre of the curve. Since every orthic curve can be brought to the above form, we see that every orthic curve is equilateral. The converse proposition, every equi- lateral is orthic, is not true. The general equation of an equilateral may be put in the form x" + ax*+ 0 (xx) =0, where @ (xx) is a perfectly general function of degree m—z2. @% con- tains $(” — 2) (~— 3) product terms, which must vanish for the curve to be orthic. To make an equilateral curve orthic is, there- fore, equivalent to making it satisfy }(“— 2) (#—3) linear con- ditions. For z==2 and = 3 this number is zero, so the equi- lateral conic and cubic are orthic. For the quartic, this says that to be orthic is one condition. Ill. W-ads, Foct, Intersections with a Circle. The relation (~—4,) (w—a,) .. . (w—a,)=pt,—2 \ 316 BROOKS—ORTHIC CURVES. [May 20: may be regarded as mapping a line through the origin in the z-plane into the orthic curve in the x-plane. The methods of analysis which were used, in the paragraphs referred to, in the study of the orthic cubic may be extended to any z, and lead to the following general theorems : On an orthic curve of order n there ts a single infinity of sets of n points, n-ads of the curve, from which all points of the curve have the same orientation. All the n-ads have the same orientation from any point of the curve (Part One, VIII). Any points may be taken as an ”-ad of an orthic curve. If we take ~ points of the unit circle as an z-ad, and find the remaining intersections of the circle and the curve, we see that they are the vertices of a regular polygon (Part One, XII). Every circle through an n-ad of an orthic curve of order n meets the curve again in the n vertices of a regular polygon. The centre of an orthic curve ts the centroid of every n-ad of the curve. For when the equation is taken in the form P@taP®Ftt+... +a w—s the origin is the centre of the curve, and is also the centroid of the m points which correspond to a point z. This equation will have two coincident roots whenever D,2 > nx +n (n—2) ax"... =0. In general, this will give 2 — 1 branch points in the z-plane. Each branch point, when reflected in the director line, gives rise to ” real foci. If'the line & revolve about a point, each reflection generates a circle (Part One, XIV). All »—1 of these circles are concentric; and they map into »—1 cassinoids, on which lie the foci of the curves which have the ”-ad which corresponds to the centre of the system of circles. These cassinoids are orthogonal trajectories of the central pencil of orthic curves. Since each of the circles must contain a branch point, each cassinoid must have at least one node. IV. Zhe Orthic Curve Referred to its Intersections with a Cirele. We know that we may put 2” linear conditions on an orthic curve. If we make it go through 2m points on the unit circle, its 1904.] BROOKS—ORTHIC CURVES. 317 equation, expressed in terms of the elementary symmetrical functions of the points where it meets the circle, becomes a — $9 4 gt 29 tg tt tg Ft — 9, The centre, found by equating to zero the » — 1" derivative as to %, iB r= ~ Sie This is the midpoint of the stroke from the centre of the circle to the centroid of the 2” points. The equation of an asymptote now takes the form H — 3 5, = "|/— Son (x Nay NERY iy) Y V. Construction of an Orthic Curve. The method which I have proposed (Part One, V) for the con- struction of an orthic cubic might be extended to the construction of any orthic curve. For this purpose the instrument must have # hands, moved by # weights. The centre of gravity of any number of weights could be held by joining them together in sets of three or less, and then joining again the centres of gravity of these sets. This operation could be repeated until the required ae of weights is reached. VI. Geometrical Characteristics. The geometrical characteristics of an orthic curve of order 1 are that it is equilateral, and that it intersects tts asymplotes in points of a second orthic curve of order n — 2. For consider the orthic curve referred to its centre, + at t—ast . .. — att a +x*—=0. The asymptotes, which are given by x ——— O°, are concurrent and equally inclined, so the curve is equilateral. The points common to the curve and its asymptotes lie on the curve a4 —agtt +... a4 ato. But this curve is of order 2 — 2, and is orthic. 318 BROOKS—ORTHIC CURVES. [May 20, To require a curve to be equilateral is to impose 2%— 3 con- ditions, and to require the curve of order 2 — 2, along which it cuts its asymptotes, to be orthic is to impose }(#" — 2) (~— 3) further conditions, in all 4% (7—1). But $2 (m—1) is the number of conditions required to make a curve of order # orthic. Part THREE—PENCILS DETERMINED BY Two ORTHIC CURVES AND ORTHOCENTRIC SETS OF POINTs. 5 I. Jntroduction. We shall now take up the study of the pencils of curves deter- mined by two orthic curves. The main purpose of this investiga- tion shall be to learn what we can about the figure of ~* points in Figure 2, The hypocycloid of class five and order six, which is enveloped by the asymptotes of curves in a pencil of orthic cubics. which two orthic curves intersect. Such a figure of m* points we shall call an Orthocentric Set, or an Orthocentric n°-point. There is a well-known proposition that all the equilateral hyper- bolas (orthic conics) which can be circumscribed to a given triangle pass through the orthocentre of the triangle. The four points, the Yertices and the orthocentre of a triangle, or, what is the same thing, the intersections of two orthic curves of the second order, have the property that the line joining any two of them is perpendicular to the line joining the other two, The term orthocentric is applied 1904.] BROOKS—ORTHIC CURVES. 319 to aset of four points related in this way. We wish to find out what metrical property distinguishes the #*-point, in which two orthic curves of order intersect. Il. Zhe Central Pencil and Jts Orthocentric Set. The first generalization which we shall make is to show that any pair of points, a, 8, together with their antipoints, a, B and 8, a, form an orthocentric four-point. a and # determine a central pencil of orthic conics, (« — a) (x —8) =1 (x—a) (x—8), and the antipoints are evidently on all the curves of the pencil. If we consider t as a parameter in the general equation of an orthic curve, (x — a) (%—a@,).. .(*—a,) =t (xn — a) (x —a,)... (x—a,), we obtain the equation of all the curves of which a, . . . a, is an n-ad. ‘The points of the orthocentric m’-point determined by this are the # real points a, and all their antipoints, But as the pencil is determined by the real points it follows that: Any n points, with all their antipoints, form a central orthocentric n*-point. The centroid of the #*-point determined in this way is the cen- troid of the # real points. The real and imaginary foci of any curve are such a set of orthocentric points. III. Zhe Pencil of Orthic Cubics through Five Points of a Circle. The Locus of Centres. We have seen that six points of a circle determine an orthic cubic curve. If the six points are 4, 4, 4, 4,4, 4, then, as we have seen, the equation of the orthic cubic through them is x? — 5,27 + 54% — 5, + 5% = 5x" + 5,x*=0. If we replace /, by a variable parameter /, and put o’s for the ele- mentary symmetrical combinations of 4, . . . 4, we have §$=%+4, 5=0,+ bo, Sg= 6, + t0,, 5,= 9; + Lo, $26, +- f4,; 5,=— 6,2. 320 BROOKS—ORTHIC CURVES. [May 20, If we make this substitution we get x*— (6, + f)x* + (4, + to,)x — (0, + toy) + (4, + t0,)x — (a, + t0,)x° + 0,02" = 0. This is the equation of a pencil of orthic cubics through five points of a circle. The centre of the curve through the six points is a= ts, If the sixth point move around the unit circle, this becomes x= $(4, + This is the map equation ofa circle. We have thus the theorem: The locus of centres of the orthic cubics through five points of a circle is a circle. Its radius ts one-third that of the given circle, and \tts centre ts the point to. M. Serret' gives an elegant synthetic proof of the theorem that the locus of centres of the curves of a pencil of equilaterals is a circle. I obtained the same result for orthic curves independently, and, as the analysis is so direct, it.seems advisable to let it stand. IV. Zhe Hypocycloid Enveloped by the Asymptotes. I shall now prove, for the pencil of orthic cubics through five points of a circle, a theorem which M. Serret' states without proof. The theorem referred to, when stated for orthic cubics of the pencil under discussion, becomes: The curve enveloped by the asymptotes of all the orthic cubics through five points of a circle is an hypocycloid of order six ana class five. It is circumscribed to the centre‘circle of the pencil, and its cusps lie on a concentric circle five times as large. We found that the equation of an asymptote, in terms of the six points where the curve cuts the unit circle, is (x — 45,) + M55(x — $557) =0. If we replace /, by the parameter /, this becomes a — (+A + Vout (@ — 4} fo; +0}) =o. ' Sur les faisceaux reguliers et les équi'atéres d’ordre #, Paul Serret, Comptes Rendus, 1895, t. 121, pp. 373-5- 1904.] BROOKS—ORTHIC CURVES. 821 we seek the curve enveloped by this line, as ¢ runs around the unit circle. For the sake of simplicity, let us refer this equation to a new system of codrdinates, so chosen that the centre circle of the pencil becomes the new unit circle. The equation becomes x—t+ Wot (x —f) =o. If now we take an axis of reals which makes ¢, = 1, and also put t® for ¢, we have att —rt «x«—r=o. The map equation of the curve enveloped by this line is obtained by equating to zero the result of differentiating with respect to +. It is x == 3r*— ar’. This is a curve of double circular motion. The curve is of order six, for it meets any line, where or 2t® — 2r® — ar’? + 37 — 3 =0. This gives six r’s, and, therefore, the curve is of the sixth order In order to determine the class of the curve, we must examine the equation of a tangent, att — 7+ ¢—rr=O0. This is of the fifth degree in the parameter, and there are, therefore, five tangents from any point x. The stationary points, or cusps, are the points where the velocity of x is zero. For sucha point we must have Drx =o, and at the same time |7| —1. Both these conditions are satisfied by T= V — =I. The curve has, therefore, five real cusps; one when r is each of the fifth roots of minus one. PROC. AMER PHILOS. SOC. XLIII. 177. U. PRINTED ocT. 19, 1904. 322 BROOKS—ORTHIC CURVES. [May 20, If we put « —— 1 we get a cusp. == 3KP 2k, x 5. e Since multiplication by «’ is equivalent to a rotation $z, we see that the locus of cusps is a circle, about the centre of the pencil, and five times as large as the centre circle. A rotation 2m sends each cusp into another, and so the cusps are equally spaced along the cusp circle. The intersections of the hypocycloid with the centre circle, xx = 1, are obtained by solving «= «* for c. We have x == 3t" — 2t*, and x= 3r?— art, The parameters of the points sought are the roots of 127° — 67” —- 6 =0, or of ( Sete cee ie: “ 22 19 185 4 125 22.6 PON Dh Dae Sah Kai « Ve 0s Sh Gao ad k dices AEBS Taking account of the different length of runs and averaging we get for runs 2 to 7— Average volts... == 17.7 “ MNDiic,. ce 163. “ efficiency = 29.1% On finally weighing the total yield of clean calcium there remained 1o50 grams. Adding 35 grams for loss by oxidation, analysis and samples (= 1085 grams), the efficiency becomes 108 29.1 X ae 26 6% If covered by an inverted graphite crucible the furnace can be left cold for more than forty-two hours, and then started again ina 386 GOODWIN—-ELECTROLYTIC CALCIUM. [Nov. 4, couple of minutes without the least trouble by drawing an arc between the iron cathode and graphite anode at the surface of the solid calcium chloride which immediately melts, allowing the iron end to be immersed and moved slowly to the centre of the furnace as the zone of fusion widens until it soon extends to the graphite on all sides. The sticks of calcium obtained were of irregular shape and covered with chloride. The bright metal showed the follow- ing composition, found by an analysis of the piece used in the tension and conductivity tests: Shs wives ths ms aree Cee b arias ke Gee 0.03% FOZ acces Gas Pes ipe Pe es cia Ox aliy 0.02 « PL rule's ogre toik poe atte ears 6 08 che: stain imiapaleoniats 0.03 §§ CBs ee i uicisd ee ee eine oy Cae ee 98.00 « Ms, isi at, Meee Peed takes a kite antic dla O.IT & CL bio wth s Weiner a Stabe sich aie’ cieh in alatene 0.90 « O (lp GEGCE a ain os ie > > > seme waa « 0.91 * 100.00 ¢« The product of run No. 5 is shown in Fig. 6. This piece was 56 cm. (22”) long, .8 cm. (35;”) least and 3.2 cm. (1%) greatest diameter, weighed 295 grams and represented a current efficiency of over 40%. Some of the other pieces were of larger diameter. The difficulty experienced until recently in making metallic calcium was probably due to the small scale on which the operation was tried. The simple and satisfactory operation of this furnace would lead one to believe that, technically, the process would be still more efficient and easily controlled. A furnace five times as large, using about 1200 amperes, would require about 8 volts, and the screw mechanism could be electrically controlled, keeping the current constant and the product perfectly uniform, as the rotary furnaces of the Union Carbide Co. are controlled. A water-cooled shield might be necessary to cool the large calcium cathode as it was drawn from the bath. The two essential conditions of operation are— 1. Rapid withdrawal of the metal formed to increase the yield and minimize recombination, 2. Narrow temperature limits. The bath must be hot enough to deposit the metal molten, not spongy, and cool enough to let it congeal upon the cathode and be raised without breaking off. To clean the metal most of the chloride was broken off with a 1904.] GOODWIN—ELECTROLYTIC CALCIUM. 387 hammer and the rest dissolved off by leaving the pieces in 95 per cent. alcohol over night. Some hydrogen was evolved but the loss due to this cause was not very great. To keep the metal for a long time it was put under oil, dipped in melted paraffin or simply put in a dry stoppered bottle. An attempt was made to fuse several pieces into one big stick. A two-foot length of one-inch iron pipe was threaded at both ends Figure 6. Figure 7. and acap screwed firmly on one end. The inside was cleaned with dilute sulphuric acid and washed with water, alcohol and ether, and in it were placed about 300 grams of clean pieces of calcium. The whole was heated in a wind furnace to a bright red heat, and on looking down into the tube one could see the red-hot molten metal which was quite fluid as shown by a thick iron wire used as a poker. The upper cap was then screwed on, the tube drawn from the fire, 388 GOODWIN—ELECTROLYTIC CALCIUM. —_[Nov. 4, its lower end hit smartly on the cement floor several times, after which the tube was quickly cooled in water. The lower cap was broken off and the walls of the tube cut lengthwise in the milling machine. When torn apart the two halves, split down the centre, displayed a most beautiful mass of large, reddish wiolet, cubical crystals (Fig. 7). There was much speculation as to the composi- tion of this peculiar ‘‘compound’’ until the following analysis showed it to be over 90 per cent. calcium: GORZ0E 6 isc ateendwes ee b een Svencs, 0.03% SOS yeaah PAE S16 044 SR TES 0.77 * PSO, temege st katint os os cas Ve Ree ee en's 0.46 *« PEO OEE Sa akclam 9's''o he's Faas 0.77 « Ca. Fired alg aaa bso «gs fae Saal 91.28 « a et Be a khakis ta sietgunta vic abe tee tens O.1E & Ch REE Pwned as bi wine's sel emanates 1,28 « Ci Fane Cine wae ees sil eo ee os trace ic otgiiet alec sale gas k's Kawls ot ckeaien nee trace O' (ESE sees cxeawae tes. 5.30 * 100.00 * The crystals showed a specific gravity of 1.5425 at 28.1° C. In water they evolved hydrogen with an oder of acetylene. Carbon was probably extracted from the iron melting tube, which reaction may be of technical importance for converting pig iron into steel, and the power of calcium to combine with and remove sulphur and phosphorus are very important as is also its strong reducing action on organic compounds, the reaction being more easily controlled and less dangerous than with metallic sodium. These crystals were quite soft and were hammered as thin as paper, often exploding with a slight flame under the impact of the hammer. When filed or cut they showed a brilliant metallic lustre, being not as pure a white as silver but slightly yellow. The solid metal at times has this same slight yellow tint. The crystals near the top of the tube evolved ammonia with water, sh wing that they had combined with the nitrogen in the melting tube. The solid metal can be worked like other metals and is much more stable than imagined. It can be heated red hot continuously in a triple Bunsen flame without igniting, but at this temperature its texture is like olay and it can be easily squeezed apart with the tongs, sometimes igniting at the edges and burning feebly till the 1904.] GOODWIN—-ELECTROLYTIC CALCIUM. 389 lime formed smothers the flame. When sent whizzing through the air against asbestos, bricks or cement it burns violently with a brilliant white light like magnesium and leaves a streak like anti- mony. It is not hardened by heating red hot and plunging into water. At 300°—400°C. it is as soft as lead and the irregular sticks can be easily hammered on an anvil, rolled, swaged or worked into any shape whatever by simply heating from time to time. When cold a bright calcium surface becomes dull rapidly in ordinary air, but if hot the metal can be brightened with a file or polished in the lathe with emery cloth and will remain bright as long as it is hot. About 300 grams of fine bright specimens were prepared as follows : A glass cylinder and its stopper were put in an air bath, gradually Figure 8. heated and kept at 150° C. The calcium was kept hot on a stove plate and the pieces polished one at a time and put in the cylinder in the air bath where they kept bright till all were polished. A - little paraffin was rubbed around the stopper and the cylinder closed. In this dry air they have lost none of their lustre and their bright surfaces are as distinctly metallic as any other metal. Fig. 8 shows the cylinders of crystals and solid metal and a six-inch stick of polished calcium. SPECIFIC GRAVITY. The density of oil was determined by the density bottle and from weighings of a bright piece of calcium in air, and in this oil its PROC. AMER. PHILOS. SOC. XLIII. 178. Z. PRINTED JAN. 3, 1905. 390 GOODWIN—ELECTROLYTIC CALCIUM. [Nov. 4, specific gravity was found to be 1.5446 at 29.2° C., which is com- pared with some other metals in the following table: SBr eet. 6.76 ORE ioees 7.00 ere 7-35 Bes ntsais 793 Ce. ess igete 8.55 , Fear 8.89 TOS scars a's ante 9.82 CONDUCTIVITY. eee e er eee eee eee wee On the milling machine a piece of calcium about 10 cm. long was accurately finished on the sides and measured 1.43 X 1.02 cm. It was imbedded in a block of wood wich a mercury cup at each end and connected through a 15-ampere Weston ammeter, variable Sharp brass potential points near the ends led to a large, very sensitive, horizontal D’Arsonval galvanometer whose deflections were read by a tele- resistance and switch with storage batteries. scope and scale. The value of the galvanometer deflections in volts was obtained by using a standard low resistance in place of the unknown piece of calcium. The.average of several readings gave the resistance between points 7.2 cm. apart 19.4 microhms at 30° C., and 26.7 microhms at 123° C. in a bath of hot paraffin. Solving the equations Rgy = Ry (1 + 30a) = 19.4 Ryos = Ry (1 + 1230) we get the resistance at o° C. Ry == 16.94 microhms and the temperature coefficient @ = .00457. Hence the specific resistance at 0° C, __ 16.94 X 1.02 K 1.43 == 3.43 microhms per cm, cube. At the mercury cups calcium slowly formed a voluminous amalgam. In the table below these values for calcium are placed in Sir Roberts-Austen’s relative electrical conductivity table and show 1904.] GOODWIN—ELECTROLYTIC CALCIUM, 391 that calcium is the fifth best conductor, being surpassed by silver, copper, gold and aluminium, if wires of equal diameter and length are compared, but for wires of equal weight and length the order is entirely different, calcium being second and exceeding silver by 67 per cent., copper by 62 per cent., gold by 86 per cent., and aluminium by almost 20 per cent. This method takes into account the specific gravity of the metals and gives the following order of conductivity : sodium, calcium, potassium, aluminium, magnesium, copper, silver, gold, etc. With purer metal still better results are to be expected. Spec. Res Relative Conductivity. 0° C. Mic- . Metal. rohms per cm. Temp. Coef, Ste tiashec| . Stovilar Cube. Wet, and Length. ry mba Be oer okacs ok 1.55 00377 100, 3265 PER Ae Ree 1.59 388 nO a one rt OG A RO 2.02 365 76.6 13. Bec cen oe 2.45 390 63. 80.4 0 EE SS SEA 3-43 457 45-1 100.0 EE sig ants bale eikisie Nery Pa Geog agape ne Eres 39.4 75-5 Michie vc abe ace Ge 5-04! 438! 31.4 IIS. MAE COR ean aie .22 365 296 14.5 (On GERACE 2 eRe ar oe ere ans sq piataianete 24.4 9-7 RE eee 7.01! 581! 22,1 86.8 PES SOARES yf gales: Palen ven eae 16.9 | 6.8 Mais thee