in a partially crystallized lava. It was evidently the last component to solidify. The composition of the rock is as follows:

SiO, AO, Fe̟O, FeO CaO MgO Na2O K,O H2O P2O Total 47.85 13.24 2.74 2.65 14.36 5.68 3.72 5.25 2.74 2.42

= 100.65 This composition is so similar to that of the Vesuvian leucitophyres, that the rock is regarded as the plutonic equivalent of these lavas. The low percentage of silica and the high lime percentage separate the rock from the eleolite syenites.

One form of the nepheline-malignite is panidiomorphic through the development of the orthoclase and all the other components in crystals. In the garnet-pyroxene phase of the rock the orthoclose is intergrown. with albite in the form of phenocrsyts imbedded in a hypidiomorphic aggregate of aegerine-augite, melanite, biotite, titanite and apatite. The augite, melanite and biotite are allotriomorphic. They seem to have crystallized contemporaneously with each other, and with a part of the orthoclase. In the amphibole-malignite the distinguishing characteristic is the prevalence of a very strongly pleochroic amphibole, and the absence of any large quantity of aegerine. The augite that is present occurs intergrown with the amphibole. Melanite is wanting, otherwise this rock is very much like the mellanite-pyroxene malignite.

The author points out the fact that the great mineralogical differences observed in the three types of malignite, are accompanied by very slight differences in chemical composition. The three types are regarded as differentiation phases of the same rock mass.

Foliated Gabbros from the Alps.-Schäfer gives an account of the olivine gabbro and its dynamically metamorphosed forms which constitute the rocks of the region in the vicinity of the Allalin glacier between the Zermatthal and the Saarthal in the Alps. The normal gabbro contains in its freshest formis much or little olivine. In its altered forms it consists of saussurite, amphibole, talc, actinolite and garnet. Ottrelite is often found enclosed in the talc and sometimes imbedded in the saussurite. In one of the granular varieties of the metamorphosed gabbro a blue amphibole is very abundant. It is intergrown in part with omphacite. The granular alteration forms of the gabbro pass gradually into foliated forms and through these into rocks called by the author “ green schists." The schistose gabbros are mineralogically similar to the granular alteration phases of the rock, except that they contain in addition to the minerals named above a 3 Neues Jahrb. f. Min., etc.. B.B., p. 91.

newly formed albite, zoisite and white mica. The final stage of the alteration is a zoisite-amphibole rock. The green schists are composed of ellipsoids of zoisite, feldspar and epidote imbedded in a schistose green amphibole clinochlor aggregate. Some of the schists are rich in garnets, and others are practically chlorite-schists. All are supposed to be derived from the gabbro.

In addition to the gabbros there are also in the region several ex. posures of serpentine whose contact with the green schists with which they are associated are always sharp. The original form of the rock is unknown, but it is supposed to have been a peridotite. Its most interesting feature is the possession of light yellow and brown crystals of some member of the humite family.

On the west side of the Matterhorn the author also found normal and olivine gabbros, both more or less altered. The former is cut by little veins of aplite. The peak of the Matterhorn is scarred by numerous fulgurites. Its rocks are fine grained green schists, some of which are like those described above, while others are dense and homogeneous in appearance. They consist of amphibole, clinochlor, zoisite, altered plagioclase, tale and alkali-mica. These rocks are defined as zoisite

amphibolites.

The Rocks of Glacier Bay, Alaska.-Cushing' gives a few additional notes on the petrography of the boulders and rocks of Glacier Bay, Alaska. The principal rocks of the region are diorites, altered argillites and limestones that are cut by dykes of igneous rocks. In addition to the diorites and quartz-diorites reported by Williams5 from this vicinity, there are also in the region mica and actinoliteschists. The dyke rocks are mainly diabases. The author gives some additional information concerning the diorites and briefly describes the schists. The actinolite schists are aggregates of finely fibrous actinolite needles, in whose interpaces is a granular mixture of quartz and epidote and an occasional grain of plagioclase. The mica schists present no unusual features except that some of them are staurolitic.

Petrographical Notes.-As long ago as 1836 Thomson reported the occurrence of light yellowish-green rounded masses which he called huronite, imbedded porphyritically in a boulder of diabase from Drummond Island. Other occurrences of the same substance have been found by the Canadian geologist in diabase dykes cutting the rocks of the Lake Huron region. These have been investigated by Barlow Trans. N. Y. Acad. Sci., Vol. XV, p. 24. Cf. AMERICAN NATURALIST, 1892, p. 698. "Ottawa Naturalist, Vol. IX, p. 25.

and are pronounced by him to be aggregates of zoisite, epidote, sericite and chlorite in a mass of basic plagioclase. In other words, huronite is a saussuritized plagioclase. Descriptions of a number of dyke rocks containing 'huronite' are given by this author.

Bauer' describes a number of specimens of snow-white, lilac and emerald-green jadeite from Thibet and upper Burmah. One of the green varieties is cut by little veins of nepheline, containing plates of basic plagioclase and little bundles of a monoclinic augite (jadeite) with the same properities as that which constitutes the mass of the jadeite. The rock, according to the author, is made up of this augite and nepheline, the latter mineral acting as a groundmass. The veins are those portions of the rock in which the augite is in very small quantity. In other specimens nepheline occurs in small quantity, and plagioclase is abundant. His conclusion is that the rock is a jadeiteplagioclose-nepheline rock in which locally the one or the other component is most prominent. If the rock is, as the author supposes, a crystalline schist, the occurrence of nepheline in it is of extreme interest.

In a second article the same authors describes a serpentine from the jadeite mines at Tauman. It is composed of olivine, picrolite, chrysotite, webskyite and a few other accessories in an albite-hornblende matrix, consisting of an aggregate of single individuals of untwinned albite, in the midst of which lie brown and gray hornblendes surrounded by zones of a bright green variety of the same mineral. Between this zone and the albite there is a fringe of green augite needles. The rocks associated with the jadeite and the serpentine are also described. Among them is a glaucophane-hornblende-schist. All the rocks exhibit the effects of pressure.

In a very short note Beck calls attention to the fact that the molecular volume of dynamically metamorphosed rocks, i. e, of the minerals composing these rocks-is less than that of the original rocks from which they are derived. For instance, a mixture of plagioclase, orthoclase and water in the proportion to form albite, zoisite, muscovite. and quartz has a molecular volume of 547.1, while the corresponding mixture of albite, zoisite, etc., has a volume of 462.5,

7 Neues Jahrb. f. Min., etc., 1896, I, p. 85.

Record of Geol. Survey of India, XXVIII, 3, 1895, p. 91. 9 Kais. Ak. Wiss. in Wien. Math. Naturw. Class, Jan., 1896.

GEOLOGY AND PALEONTOLOGY.

Phylogeny of the Dipnoi.-In a memoir recently published, M. Dollo adduces fresh evidence for the theory recently advanced by various English scientists that the diphycercy among the Dipnoi is in reality a secondary diphycercy or gephyrocercy. The author regards this as an important fact, and uses it as a basis in developing his theory of the origin and evolution of the order. The results of his researches are as follows:

I. Dipterus valenciennesii is the most primitive of the Dipnoi known. II. In a general way, the evolution of Dipnoi, since the lower Devonian, is represented in the following series in the order of the enumeration of its terms.

Dipterus valenciennesii-Dipterus macropterus-Scaumenacia-Phan

eropleuron-Uronemus-Ctenodus-Ceratodus-Protopterus-Lepido

siren.

III. The origin of the Dipnoi must be looked for among the Crossopterygia.

IV. The Batrachians are not in the line of the Dipnoi.

V. The specialization of the Dipnoi has been from a pisciform type toward an anguilliform type.

This order is a terminal group, derived from the stem that gave origin to the Batrachians.

In conclusion the author gives the phylogeny of the gnathostome vertebrates in a tabulated form. (Bull. Soc. Belge de Geol. T. ix, Fasc. 1, 1895, Bruxelles, 1896).

Fauna of the Knoxville Beds.-The Knoxville beds are a Cretaceous series confined to the coast ranges of California, Oregon and Washington. They are characterized by the great abundance of Aucella, usually without associates, but, through the explorations of Mr. Diller and other geologists, a rich and varied invertebrate fauna has been discovered in the Aucella-bearing series of the Pacific States. The description of this fauna was assigned to T. W. Stanton, and is now published as Bulletin No. 133 of the U. S. Geological Survey. The author recognizes 77 distinct species and varieties, of which 50 are All but 7 of the species are Mollusca, including 33 species of

new.

1 Bulletin United States Geological Survey, No. 113. Contributions to the Cretaceous Paleontology of the Pacific Coast; The Fauna of the Knoxville Beds. By T. W. Stanton. Washington, 1895, [issued Feb. 3, 1896].

Pelecypoda, 1 of Scaphopoda, 18 of Gastropoda, 18 of Cephalopoda (15 Ammonoids, 3 Belemnites). The other 7 species include 5 Brachiopoda and 2 Echinodermata.

The brief introduction comprises a geological description of the beds, with a discussion of their age, and of their relations to various other formations characterized by a similar fauna.

The new species are figured on twenty page plates.

Notes on the Fossil Mammalia of Europe, IV.-ON THE PSEUDOEQUINES OF THE UPPER EOCENE OF FRANCE.-Under the term Pseudoequines may be included the various species of the genus Paloplotherium, which occur in the Upper Eocene and Oligocene of Europe. This phylum parallels in a remarkable manner many of the characters which are typical of the true horses, but these characters, strange to say, are much earlier differentiated than in the real equine phylum.

Kowalevsky in his great work on "Anthracotherium" "Anthracotherium" clearly recognizes in his phylogenetic table of the Ungulates, that Paloplotherium is not in the direct line of the horses. Schlosser is also of the same opinion as Kowalevsky in regard to the relations of Paloplotherium to the horses. Professor Gaudry as late as 1888 placed all the species of Paloplotherium in the direct line leading to Equus.

The earliest known species referred to Paloplotherium is the P. codiciense Gaudry; this form is from the Calcaire Grossier or Middle Eocene. In P. codiciense there are four upper and lower premolars, whereas in the more typical species of Paloplotherium from later deposits there are only three premolars. Moreover, in the P. codiciense all the upper premolars are simpler in structure than the true molars. The last upper premolar in this species is tritubucular in structure, and there are two well defined crests running outwards from the deuterocone, in other words this tooth is well adapted for further evolution into the molariform last premolar of the typical Paloplotheroids. In the true molars of P. codiciense the ectoloph has nearly the same form as that of the Palaeotheridæ in general, the metaloph or posterior crest, however, is less oblique in position than in the later species of Paloplotherium and Palæotherium. The type specimen of Paloplotherium codiciense consists of a facial portion of a skull with the teeth well preserved. This species is much larger than P. minus and corresponds more nearly in size with P. annectens.