At the close of the Glacial epoch the moraine matter was reassorted into marine deposits, which in this country have been exposed to a general and sweeping denudation. Only small patches are found remaining in sheltered positions. These marine deposits consist of finely laminated clays resting upon coarser, more stony, and gravelly beds. The former were evidently estuary deposits, the latter thrown down in deeper water, where the strong Arctic current prevailed. The oldest beds are the coarser strata, which, as in Maine, occur at high-tide mark. The more recent beds occur from ten to twenty feet above the sea level.

The fossil Invertebrata, found abundantly in these beds, afford excellent material for comparison with the present marine fauna of Labrador, and throw new light on the distribution of marine life during the close of the Glacial epoch. The assemblage is thoroughly Arctic in character, but, when compared with lists of the glacial shells of the north of Europe, it is found to bear a very distinct facies. It is evident that on each side of the Atlantic, the same faunal distinctions obtained during this period as now. There was, however, a greater range in space of purely Arctic species, and, though the European marine fauna was much more closely allied to our own, owing to the great predominance of exclusively Arctic forms, it is yet evident that the Arctic glacial fauna was divided into a Scandinavian district, and a Labrador district, each the metropolis of a small number of species peculiar to itself and limited to its area.

The assemblages found at various points along the coast from Labrador to Maine are not the exact equivalents of the present fauna. They differ in containing a very small percentage of extinct species, and in a different grouping of species still living. Thus, in the Labrador beds are several species of Fusus (Sipho) which differ from recent Arctic forms, and also a species of Bela; certain forms, such as Panopea and perhaps Cyrtodaria, which were abundant formerly, seem to be dying out at the present day. In Maine the change is still more marked. Thus, the most characteristic shell of the marine clays is Leda truncata (Portlandica), which has wholly disappeared from the seas south of the circumpolar regions, unless future deep-sea dredging reveals its presence in some of the abysses off our coast. An undescribed Macoma is also characteristic of the beds about Portland; and other important changes have occurred in the relative abundance of species, and the manner in which they are grouped as compared with the present assemblages in zoological districts farther north, and similar in physical surroundings to the glacial seas.

The Labrador district of the Arctic fauna, instead of being restricted as now to the eastern coast of North America from the Arctic archipelago to the banks of Newfoundland, and shading

off into the Acadian district at the present line of floating ice, during the Glacial epoch extended up the St. Lawrence river, and as far as Portland, on the coast of Maine, where it shaded into a more southern assemblage.

In Maine there are two distinct horizons of life. The lowest and oldest is found at the bottom of the boulder-clay at hightide mark along the coast. The second horizon is composed of rewashed, finely laminated, less stony clays occupying the coast from 25 feet above the sea level to a height, 50 to 100 miles inland, of nearly 300 feet. The species found in this second horizon are rather boreal forms than purely Arctic. In the beds about Saco and Scarboro' we find Leda tenuisulcata intermingled with the Arctic L. pernula, as it is not at present on the coast, and Pandora trilineata replaces the Arctic Pandorina arenosa. At Berwick, Astarte castanea, a boreal form, is introduced; while south of this, at Point Shirley, Desor and Stimpson found Nassa trivittata, Buccinum plicosum, Astarte castanea and Venus mercenaria, species which now, as an assemblage, abound most on the shores of New England south of Cape Cod, and in New York bay. Again, at Nantucket, Desor found a still warmer fauna occupying, apparently, an extension of this second horizon. Arca transversa, Crepidula fornicata, with Buccinum plicosum and Nassa obsoleta were found to abound in this locality, where the warming influence of the Gulf Stream was strongly felt, while the waters of Maine were cooled down by the Arctic or Polar current.

In the beds of this horizon at Gardiner occur the teeth of the bison, walrus, and bones of other animals, and the Mallotus villosus; also in the same beds at Bangor the fossil whale, and in Burlington, Vt., in the Champlain clays, which evidently belong to this horizon, the Beluga Vermontana of Thompson.

Thus the two glacial fauna that have successively gained a foothold in northeastern temperate America, seemed, as regards both their land and marine animals, and also plants, (for Potentilla tridentata which is found only in Maine, Labrador and Greenland, is also found fossil in the Ottowa clays, according to Dr. Dawson,) to be a purely Arctic American assemblage. According to the view of Dr. Hooker,' the most ancient glacial flora was derived from Scandinavia. On the contrary, as far as geological evidence at present tends, the cave mammals of Europe were associated with the musk ox, reindeer, white bear, and other Arctic animals which abound in Arctic America, while no features in the Post-tertiary fossils of America seem to be EuroThese faunal distinctions would seem to be even more pean. strongly marked than now in the distribution of the Vertebrata during the closing part of the Glacial epoch.

2 Outlines of the Distribution of Arctic plants, by J. D. Hooker, M.D. Trans. Linn. Soc. London, xxiii, part ii.

ART. VII.-On a Process of Elementary Analysis admitting of the determination of Carbon, Hydrogen and Nitrogen at a single combustion; by C. GILBERT WHEELER.

THE processes of ultimate organic analysis heretofore employ. ed, when applied to substances containing carbon, hydrogen and nitrogen, contemplate the determination of the latter element by a separate and distinct operation and are therefore often embarrassing where the chemist has but a small amount of the substance to be analyzed at his disposal. A method by which these three elements might be determined at a single combustion would appear to be desirable, and, if equally accurate with other approved processes, and also admitting of a not less expeditious execution, might with advantage be employed, not only in the special class of cases referred to, but also in general, as a substitute for those now in use. For some time past I have made use in my laboratory of a method of analysis which apparently is of universal application when carbon, hydrogen and nitrogen are to be determined, and which I can recommend as yielding quite satisfactory results.

The method in question may be considered as a combination, with necessary modifications, of that ordinarily used in deter mining carbon and hydrogen with Simpson's method of determining nitrogen, as will be seen from the description which I now proceed to give, first in outline, then more in detail.

The operation commences with expelling the air from the combustion tube, by means of a stream of oxygen gas. As in the anterior portion of the tube there is placed a quantity of metallic copper, to prevent the formation of deutoxyd of nitrogen, the former cannot be heated without the copper oxydizing and the tube again becoming wholly or partially filled with air. This is prevented by expelling in turn the oxygen by means of a known quantity of carbonic acid gas, which effected, the combustion proper is commenced and carried on in the ordinary manner with this modification, viz: that, at the close, a current of oxygen gas is again employed for the purpose of forcing into the appropriate apparatus the residuary products of combustion, as also, when necessary to complete the oxydation of particles of the substance yet unconsumed. The water and carbonic acid resulting from the analysis are absorbed in the usual apparatus and weighed, while the nitrogen, mixed with oxygen, is conducted into a special apparatus where it is measured over mer

cury.

The simplest method would be to conduct the gases into a Bunsen's eudiometer, and in the same perform the necessary AM. JOUR. SCI.-SECOND SERIES, VOL. XLI, No. 121.-JAN., 1866.

further determinations. However, the capacity of such a eudiometer is not sufficient in the great majority of cases, even where it has a length of 800 to 1000 millimeters. I have therefore made use of Bunsen's gasometer' and found the same exceedingly well adapted for the purpose. It has, however, proved desirable to make one or two trifling alterations which in practice I have found to enhance its value and usefulness. Instead of the original arrangement for regulating the escape of gas, I make use of a screw-slip as shown at g, fig. 2. Also the tube cg, which is not capillary but somewhat larger, is filled with mercury and thus remains during the combustion. The filling is conveniently effected by pouring mercury into the gasometer through the tube e, fig. 3, until it passes the point g, fig. 2, and flows out in a constant, regular stream at e, which is closed with a piece of soft wax. The screw-slip being closed, the gasometer is ready for use.

Both the gasometer above described and the eudiometer, in which the mixture of nitrogen and oxygen is to be analyzed, are graduated, and their cubic capacities referred to a common standard (cubic centimeter). This is necessary, as only an aliquot part of the total gas obtained is submitted to analysis, the greater portion being held in reserve in case any disaster should occur to that under investigation.

I pass now to a detailed description of the mode of procedure in performing an elementary analysis by the process under consideration.

A combustion tube 2-23 feet in length is sealed up at one end, as shown in fig. 1. From 3 to 5 grams pure, well-dried and pulverized chlorate of potassa are introduced and thoroughly mixed with the aid of a mixing wire, with at least an equal * Bunsen, Gasom. Methoden, p. 21, fig. 16.

volume of freshly ignited oxyd of copper. It is important that the mixture be as uniform as possible in order that the supply of oxygen may be easily regulated. Then follows about two inches of oxyd of copper, and thereupon a weighed portion, from 0-2 to 03 gram, of oxalate of lead which is likewise thoroughly mixed with oxyd of copper. This salt is the substance I find the most convenient from which to obtain a known amount of carbonic acid gas.

I adopt oxalate of lead in preference to other substances that have hitherto been employed in organic analysis for evolving carbonic acid, as more completely free from the various practical objections presented by the latter, and in particular by the following, viz: carbonate of magnesia (magnesite), carbonate of manganese, carbonate of copper, bicarbonate of soda and oxalic acid. The use of these substances is impracticable, as they either fail to give the theoretical percentage of carbonic acid on being heated, partially decompose when exposed to the air in a moist condition, yield also water with carbonic acid, are too hygroscopic, or present other difficulties that render their use in this process inexpedient. Carbonate of lead is less objectionable than the above mentioned, and it is mainly on account of the greater amount of carbonic acid furnished by the oxalate that I prefer it.

I prepare the latter salt by adding to a solution of acetate of lead a slight excess of oxalic acid, and thoroughly wash the precipitate obtained by decantation. It has the formula PbO, C,O,, and yields on gentle ignition with oxyd of copper precisely two equivalents of carbonic acid. As a mean of several nearly identical results I obtained 29.83 p. c. of carbonic acid instead of 29.81 as required by theory. A greater amount of this salt than 0.3 gram being never used, the maximum error, therefore, possible in a carbon determination would be 0-00001643 gram.

After introducing, as previously explained, the oxalate of lead into the combustion tube, about two inches of pure oxyd of copper are added, then a mixture of the substance for analysis with oxyd of copper, and thereupon, again, several inches of the oxyd. Finally, the remaining space in the tube, which should never be less than 4 inches, and need not exceed, except in rare cases, 8 inches, is filled with metallic copper freshly reduced in a stream of hydrogen gas. I prefer for the purpose a compact roll of copper wire gauze.

After a channel has been secured in the usual manner throughout the whole length of the tube-at the posterior end it is well to have the channel larger than elsewhere--a chlorid of calcium. tube and potash bulbs are placed in communication as in an ordinary combustion and the anterior end of the latter is connected