only because the recent death of Tchébychev, followed in less than two months by that of Cayley, gives them now a special pertinence, but because it is of interest to compare one with what is given on 'tram motion' in Kempe's 'How to Draw a Straight Line,' and the other with its reproduction by no less a master than Clifford on pages 149, 150 of his Dynamic, whence I add figure 2.

"Robert's theorem of 3-bar motion takes the following elegant form: Take a triangle

B

B1

C

C1 ABC and a point O and through O draw lines parallel to the sides as in the figure, the 3 shaded A's are of course similar to ABC. Now imagine a linkage composed of the shaded A's and the bars AA2, AA3, BB3, BB1, CC1, CC, pivoted together at A, B, C, A2, A3, B3, B1, C2, C1, O; then, however, the figure is moved [of course A3, B3 do not continue in the line AB, etc.], the triangle ABC will remain similar to the shaded triangles; and if in any position of the figure we fix the points A, B, C, then the point O will be movable in a curve, viz.: we have the same curve described by O considered as the vertex of OA, B,, where the two radii are AA3, BB3-by O considered as the vertex of OA, C, etc.-and by O considered as the vertex of OB, C,, etc."

CAMBRIDGE, Feb. 22, 1876.

"The porism is very pretty; it was new to me, though I think it ought not to have been so. Look at the theorem thus: Imagine a plane, two points thereof, A, C moving in fixed lines Ox, Oy. Describe the circle OAC, which consider as a circle fixed

THE PROTOLENUS FAUNA.* THE above article will be one of especial interest to students of the early Paleozoic faunas, since it describes one of the oldest known.

From time to time during the last thirty or forty years discoveries of fossils have been made in the Cambrian rocks of eastern Canada. Those of the St. Lawrence valley and northern Newfoundland were by Billings referred to the 'Lower Potsdam,' but at a later date, together with others found in that valley and in southern Newfoundland, they have been more specially correlated with the Olenellus Fauna by C. D. Walcott and others.

Other fossils found in the lower part of the Cambrian rocks in New Brunswick below the Paradoxides bed were naturally at first thought to be also of this fauna, but, as will be seen by considerations advanced further on, it does not now seem possible so to establish the relationship.

The discoveries in New Brunswick have from time to time been reported in articles published by G. F. Matthew in the Transactions of the Royal Society of Canada, but such important additions were disclosed through the collections made by W. D. Matthew in 1892 and 1893, and by him in conjunction with G. van Ingen for Columbia College, New York, in 1894, that a special article on this, the Protolenus fauna, has been written. From this article the following abstract has been made of the character of the fauna, and the conclusions arrived at from its study.

The fauna consists of Foraminifera, Sponges, Molluscs and Crustaceans. All the Foraminifera described are referred to the

genera Orbulina and Globigerina; the sponges include Protospongia and others. The molluscs are mostly hyalithoid shells

*Abstract of a paper communicated to the New York Academy of Sciences by G. F. Matthew, of St. John, N. B.

of the genera Orthotheca, Hyolithus and Diplotheca. A remarkable mollusc having a helicoid shell and supposed to be a Heteropod, enables me to establish a new genus. The Crustaceans are chiefly of two groups, Ostracoda and Trilobita, of which the former are remarkable for the large number of genera and species, as compared with the trilobites; two predominant and characteristic genera are Hipponicharion and Beyrichona. All the trilobites are of genera peculiar to this fauna, except Ellipsocephalus, which, although one of the dominating types, also cccurs in the Paradoxides beds of Europe. The most characteristic genus or trilobites is Protolenus, which is abundantly present in the typical beds.

The following are some of the salient characters of the fauna as at present known. All the trilobites have continuous eyelobes. This is a decidedly primitive character, and its value in this respect is shown by the genus Paradoxides of the overlying fauna, which began with small species having such eyelobes, and culminated in the large forms of the upper Paradoxides beds in which the eye-lobe was considerably shortened. This shortening of the eyelobe was carried still further in the Oleni of the Upper Cambrian, dwarfed forms, with a general similarity to the Paradoxides, in which the eyelobe is almost on a line with the front of the glabella.

The important family of Ptychoparidæ is absent. This family did not have continuous eyelobes, for in the young, when this projecting fold first shows itself, it is short and at the lateral margin of the head-shield. No trilobite with such an eyelobe has been found in this fauna. The Ptychoparida had about a dozen species in the Olenellus Fauna, and became quite common in that with Paradoxides, and continued to abound throughout the Cambrian period.

The genus Conocoryphe is absent. This is specially a type of the Lower Paradoxides

beds and under, the name of Conocoryphe trilineata (Atops trilineatus), is claimed as a characteristic fossil of the Olenellus Zone.

The genus Microdiscus is absent. This trilobite is especially characteristic of the Olenellus Zone and continued to live with Paradoxides. Here it occurs in the Paradoxides Zone, but is absent from the Protolenus Fauna.

The genus Olenellus is absent. Though carefully looked for, no example of this genus has been found among the trilobites of the Protolenus Fauna, hence, though this fauna apparently holds the place where we might naturally expect to find Olenellus, that genus proves to be absent, or at least not at all characteristic; and, as so many of its associate genera also are absent, we cannot regard this fauna as the Fauna of Olenellus.

Of the genera of trilobites that are present Micmacca has affinity with Zacanthoides. It differs in the course of the posterior exterior of the dorsal suture. The relation

will seem closer if we suppose a movement of the eyelobe during the growth of Zacanthoides similar to that which occurred in the Ptychoparidæ, by which the eyelobe was drawn in toward the glabella, while at the same time there was a projection of the posterior extension of the dorsal suture outward toward the general angle. If this change were shown to have occurred in Zacanthoides, Micmacca might be looked upon as an ancestral form of that genus.

In this fauna there is a very primitive assemblage of Brachiopods, of forms which it is in many cases difficult to assign to any known genus. Many are small, some are minute, and the larger species belong to the Obolidæ and Siphonotretidæ.

The Gasteropoda have already been alluded to; among these Pelagiella (n. gen.) is remarkable for the peculiar aperture which seems to indicate a free swimming Heteropod.

This fauna is distinguished from that of

Olenellus by two marked features; it is more primitive and also more pelagic.

The way in which the trilobites are bound together by the single feature of a continuous eyelobe shows a unity of origin and a close relationship not found in any other fauna. And yet among these trilobites there are forms which in other respects are parallel to the types which developed in the later faunas; thus in Protolenus we have have the flat pleura with the diagonal furrow of Paradoxides and the deeply grooved, geniculate pleura of Ptychoparia, and at the same time the prominent glabella and deep dorsal furrows of Solenopleura. Micmacca, as has already been said predicated Zacanthoides of a later fauna, and Protagraulos in its almost obliterated glabella and flat cephalic shield closely resembles Agraulos of the Paradoxides Fauna.

It is a more pelagic fauna than that of Olenellus, for we notice the absence of many forms differentiated for shore-conditions. Trilobites with fixed outer cheeks, like Olenellus and Microdiscus are absent; calcareous corals and sponges are rare; thickshelled brachiopods and the Orthidæ are wanting, or rare; no Lamellibranch is known, but Foraminifera are quite common in some of the beds.

The question of the antiquity of this fauna as compared with that of Olenellus is discussed. The facies of the fauna as above described indicates a greater antiquity, but if the two faunas were contemporaneous, that of Olenellus may have reached these shores first.

VOLCANIC DUST IN TEXAS.

SOMETIME since the writer was given, for examination by the microscope, a sample of a white, fine-grained silicious deposit by Prof. R. T. Hill, of the U. S. Geological Survey, who writes as follows concerning it :

"The material which I gave you was collected by an old Texas friend of mine, Mr. S. P. Ford, in De

cember, 1893, who said that at first he supposed it was chalk, but had since come to the conclusion that it is something else. When I wrote to Mr. Ford that I thought it was volcanic glass, probably derived from some of the now extinct vents along the Rocky Mountain front, he expressed some doubt as to this mode of origin, and said:

"This specimen was from a solid hill from thirty to forty feet high, composed entirely of this stuff. The point I make is that, on account of its thickness, the crater must have been somewhere very close, and if so, is it not something heretofore unknown in Texas? The exact locality is on Duck creek, in Dickens county, about 50 miles northwest of the Double Mountain.' (Dickens county is in northwestern Texas, in the Brazos River drainage.—Author.)

"This specimen undoubtedly comes from the postCretaceous formations constituting the great Llano Estacado. Perhaps you will remember that in 1886 I collected some similar material from near Wray, Colorado, and Hecla, Nebraska, which was described by Prof. Merrill of the National Museum, in the American Journal of Science. This Texas material seems very similar to that of the Colorada-Nebraska locality, both in appearance and in geological position. I wish that more was known of the stratigraphy of the Texas beds. The Colorado specimens occur in what is called the White River Tertiary."

An examination by the microscope shows that the white material is volcanic glass, in the angular and fluted forms figured by Merrill, as characteristic of volcanic dust from Furnas county, in southern Nebraska. Diller† also describes and figures similar forms of glass particles from Norway, Krakatoa, Truckee River and Breakhart Hill, the latter a hill to the north of Boston, Mass. In the same article he describes volcanic dust from Unalashka, which fell in October, 1883, and discusses volcanic dusts in general. Professor Diller concludes that 66 so far as definite observations have been made, they warrant the general assertion, that with occasional exceptions, which can be readily explained, volanic dust contains a higher percentage of silica than the lava to which it belongs."

Professor Diller has also described some

*Proc. U. S. Nat. Mus. 1885, p. 100. + SCIENCE, May 30, 1884.

volcanic material from Knox county, Nebraska, and from the West Blue River, Seward county, Nebraska,* and estimated that about 90% was vocanic dust, there being also numerous rolled quartz grains.

The description of the material collected by Professor Hill from Wray (B. & L. R. R.), on the south side of the Republican River, occurs in an interesting article by Professor Merrill, On the Composition of Certain Pliocene Sandstones from Montana and Idaho.'t

Three figures are given showing the shape of the particles of volcanic glass found in the sandstones. In the material from the Devil's Pathway (No. 35893) "there are many disc-like bodies on the glass particles, colorless and nearly circular in outline," but the other figures show angular and fluted forms like those above referred to. Merrill gives analyses of three samples of the volcanic dust from Montana and Idaho, and concludes that they are of andesitic or tractytic origin. His analyses include lime and alkali determinations, and the silica contents range from 67.76% to 68.92%.

Merrill also states that some volcanic dust from Krakatoa fell on a ship 885 miles from the source of volcanic activity, so that the existence of a layer of volcanic dust at a given point may not indicate the proximity of the volcano from which the material came, but a deposit forty or more feet thick would hardly form at a great distance from the source.

The volcanic dust obtained by the writer from a layer in the Neocene Lake beds that underlie Mohawk Valley, in Plumas county, California, likewise resembles in the shape of its particles the dusts figured by Diller and Merrill. An analysis of this material by Dr. W. H. Melville showed that it contained 70.64% of silica, and it was there*See article by J. E. Todd, SCIENCE, Vol. VII., p. 373.

† Am. Jour. Sci., Vol. XXXII., pp. 199–204.

fore presumed to be a rhyolitic glass. The material obtained by Professor H: 11 closely resembles the Mohawk Valley material. The Texas occurrence is of unusual interest, being in a region where evidences of the former existence of volcanoes are rare. H. W. TURNER.

WASHINGTON.

CURRENT NOTES ON ANTHROPOLOGY (VI.). THE CAUCASIC LINGUISTIC STOCK.

COL. R. VON ERCKERT, of the Russian army, already known for an excellent work on the ethnography of the Caucasus, has just published an epoch-making volume on the languages of that region (Die Sprachen des Kaukasischen Stammes, Vienna, 1895). In this he solves the intricate problem which has so long puzzled linguists as to the relationship and place of these tongues. He demonstrates by satisfactory evidence, structural and lexicographical, that these numerous languages and dialects, some thirty in number (the Ossetic, which is Aryan, being of course excluded), belong to one family, which should be called the 'Caucasic.' It is divided in three groups, the Georgian, the Circassian and the Lesghian. stock stands wholly independent, all similarities to either Ural-Altaic or Indo-European proving accidental or unimportant. Which of the groups is nearest the ancient original tongue he does not pretend to decide; but he offers striking testimony to the persistence of the traits of these languages. The Georgian was written as early as the ninth century A. D., and he gives a letter composed by a bishop in 918. It is quite identical, both in syntax and words, with the current tongue of to-day.

The

All these facts are the more to the purpose since so much has been made of late years by Professors Sayce, Hommell and their followers, of what they call the 'Ala

* Bull. Phil. Soc. Washington, Vol. XI., p. 389.

CUNEIFORM INSCRIPTIONS.

DR. HUGO WINCKLER, in his 'History of Babylonia and Assyria,' tells us that the cuneiform method of writing was in use among eight nations speaking entirely different languages. Whether this is quite accurate or not, we need not stop to consider, as there can be no question that it had a much wider distribution than used to be supposed. Last year the well-known French archæologist, M. E. Chantre, unearthed specimens of it at Pterium and Cæsarea, in Asia Minor, as far west, perhaps, as such inscriptions have been found in place. The excavations continued by the University of Pennsylvania at Niffer have proved rich in finds of tablets. But the champion recent discoveries appear to be those of M. de Sarzec at Tello. A brief account of his eighth campaign in that rich locality appears in the Révue Archaeologique' of December last, extracted from the official report of M. S. Reinach. From it we learn that M. de Sarzec opened a small mound some hundreds of yards from that which he had previously worked, and chanced upon the very archives of the old city themselves. They were inscribed on tablets and neatly stored in trenches, where they had rested undisturbed these thousands of years. From these deposits he took out more than thirty thousand tablets, about five thousand in perfect condition, another five thousand very slightly injured, and the others more or less defaced. This magnificent discovery will have the greatest importance in revealing the history and character of the ancient Babylonian civilization.