consisting of hundreds of small spicules, makes the solution of them possible in a very short time.

The gastroliths have been supposed to possess great medical properties and to perform a variety of functions, the most common and accepted belief being that they play an important part in the provision of lime for the hardening of the new shell. The small quantity of lime which they contain, however, not more than one one hundred and twenty-sixth of that of the entire shell, according to an analysis recently made by Dr. Robt. Irvine, shows that this is relatively unimportant. Fragments of lime furthermore are always at hand, and are frequently eaten by the soft lobster, shortly after ecdysis, in the adolescent stages at least. It is more likely that the gastroliths are the result of excretion of lime which is absorbed from parts of the shell to render molting possible, and that their subsequent absorption in the stomach is a matter of minor importance.

Rate of Growth.-Larvæ increase in length at each molt (stages 2 to 10) from 11 to 15.84%, or on the average about 13.5% (measurements from 66 individuals). The increase in the young at each molt agrees quite closely with that seen in the adult, where the increase per cent. in ten cases was 15.3%. Allowing an increase per cent. at each molt of 15.3-probably not excessive for young reared in the ocean-and assuming the length of the first larvæ to be 7.84 mm. we can compute approximately the length of the individual at each molt.

times; and a 19-inch lobster, 30 times. These estimates do not, I believe, go very far astray. We see them practically verified up to the tenth molt.

The time interval between successive molts is the next point to consider. Here the data are very imperfect. How long is the three-inch lobster in growing to be six inches long? Probably not more than two years and possibly less. This is supported by the observations of G. Brook. We therefore conclude that a ten-inch lobster is between four and five years old, with the highest degree of probability in favor of the smaller number.

ADELBERT COLLEGE.

FRANCIS H. HERRICK.

THE NEWARK SYSTEM.

IN an article in a recent number of SCIENCE* Professor C. H. Hitchcock again objects to the use of Newark' as a group name in geology. This article is essentially a republication of a portion of a paper by the same author, which appeared in the American Geologist in 1890† in criticism of an article of mine in the same journal,‡ in which reasons were presented for reviving the use of Newark as a name for a certain system of rocks.

I replied to Professor Hitchcock's objections and criticisms, and showed conclusively, as I believe, that the term referred to has precedence over all other names applied to the system in question, which do not imply correlation. In his recent article Professor Hitchcock does not so much as mention my rejoinder; but is of the opinion that the considerations presented in his earlier paper would have been sufficient to convince any one, looking at the subject judicially and impartially, of the inadequacy

* Vol. 1, New Series, Jan. 18, 1895, pp. 74-77.
† Vol. 5, 1890, pp. 197-202.
Vol. 3, 1889, pp. 178-182.
Am. Geol., Vol. 7, 1891, pp. 238-241.

of the name Newark to special recognition.' On the other hand, I am of the opinion that my reply should have silenced opposition. There is, thus, a radical difference of opinion between us. There is also a question of fact involved. Has Newark priority as a group name? This is a simple historical question that almost any one can decide from the documentary evidence. In the papers described in the following foot-note* I have presented or referred to all of the evidence known to me bearing on the question.

In Professor Hitchcock's recent article there are many statements that have no relation to the matter under discussion, since they refer to usages of later date than the introduction of the term Newark. No legitimate arguments are advanced that are not in the former paper, and as these have all been answered, there is nothing left for me to do but to follow my opponent's example and republish my reply to his five-year-old criticism.

lationship with distant formations was also pointed out. The first name on the list referred to which met this requirement was Newark group,' proposed by W. C. Redfield, in 1856. That this was a group name, intended to indicate the entire formation, is shown by the language used. Redfield's words are:

"I propose the latter designation [Newark group] as a convenient name for these rocks (the red sandstone extending from New Jersey to Virginia) and to those of the Connecticut valley, with which they are thoroughly identified by foot-prints and other fossils, and I would include also the contemporaneous sandstones of Virginia and North Carolina."*

"IN a brief paper on the Newark system published in this journal [Am. Geol.] about two years since,† I proposed a revival of 'Newark' as a group name for the reddishbrown sandstones and shales and associated trap rocks of the Atlantic coast region, which had previously been quite generally referred to the Triassic and Jurassic. A long list of names was presented that had been used to designate the rocks in question; nearly all of which implied correlation with European terranes, ranging from the Silurian to the Jurassic. The advisability of adopting a name that did not indicate re

*The Newark System, Am. Geol., Vol. 3, 1889, pp. 178-182.

Has 'Newark' priority as a group name, Am. Geol., Vol. 7, 1891, pp. 238-241.

The Newark System, U. S. Geol. Surv., Bull. No.

85 (Correlation Papers) 1892.

† Vol. 3, 1889, pp. 178–182.

† Am. Geol. 5, April, 1889, p. 251.

been an exhaustive examination of the literature relating to the terrane in question, and concluded that Redfield's name had precedence over all other names that had been used which did not imply correlation.

The term Newark system has recently been adopted by several geologists, in accordance with my suggestion, and up to the present time but one voice has been raised against it. In an article on 'The use of the terms Laurentian and Newark in geological treatises,' published in this journal, † Prof. C. H. Hitchcock has formulated five objections to its acceptance. These will be considered in the order in which they were presented.

First. It is claimed that 'An essential feature of a name derived from a geograph

* Am. Jour. Sci., 2d ser. 1856, Vol. 22, p. 357 ; also in Am. Assoc. Adv. Sci., Proc., Vol. 10, Albany meeting, 1856, p. 181.

† Vol. 5, 1890, pp. 197-202.

ical locality is that the terrane should be exhibited there in its entirety or maximum development;' and that the territory about Newark, N. J., does not meet these requirements for the Newark system.

Without dissenting from the wisdom of the rule proposed, although a large number of exceptions could be found to it in the best geological memoirs, I wish to state from my own knowledge that the region about Newark may be taken as typical of the terrane named after that city. The characteristic reddish-brown standstones and shales are there well exposed, and in the neighboring Newark mountains the associated trap rock occurs in sheets of great thickness. This statement is sustained by Prof. Hitchcock's own words, a little farther on in the paper cited, where he says, "the New Jersey terrane possesses the distinguishing features of the Trias quite as well as the one in New England."

That Passaic would have been a better name, as Prof. Hitchcock suggests, is perhaps true, but the one before us was definitely selected and has priority.

Second. It is stated by Prof. Hitchcock that the name 'Connecticut or Connecticut River sandstone has priority over Newark,' and was used by several geologists before Redfield's proposal in 1856, though none of them had proposed it as a geological term.' The admitted fact that no one had used the name referred to as a geological term, relieves me of the necessity of showing that Redfield's name has priority.

In the writings of the older geologists among whom Prof. Edward Hitchcock will always take the first rank as an investigator of the sandstones of the Connecticut valley, the terms 'Connecticut sandstone,' or 'Connecticut River sandstone,' were used in the same sense as the coördinate term I have just employed, i. e., as a geographical designation; just as they might have referred to the granite of Massachusetts without any

intention of proposing a group name. The fact that the older geologists, and among them Prof. Edward Hitchcock, spoke of the Newark rocks of New England under definite group names, implying correlation, is sufficient evidence that they did not recognize the value of an independent name.

Third. It is stated that Prof. J. D. Dana adopted the name proposed by Redfield, in his lectures, but did not use it in his subsequent writings. Prof. Dana's reasons for this course have never been published, and so far as it is a precedent-happily precedents have less weight in geology than in some other professions-it indicates that we should first use the name Newark and then abandon it for other names implying indefinite correlation with distant terranes.

Fourth and Fifth. While it is admitted that the terrane under discussion is quite as well represented in New Jersey as in the Connecticut valley, it is claimed that the latter having been studied first, should have furnished the group name. I fully agree with Prof. Hitchcock in this, and could add several other group names which to my taste might be improved, but the author of a geological name, like the paleontologist who describes a new fossil, is entitled to priority. To attempt to introduce a new name for a group of rocks already sufficiently well designated, would only bring confusion, similar to that produced by the great variety of names implying correlation that have already been used for the Newark system." ISRAEL C. RUSSELL.

UNIVERSITY OF MICHIGAN.

DEATH OF GEORGE N. LAWRENCE.

THE veteran ornithologist, George N. Lawrence, died at his home in New York City, Jan. 17, 1895, at the age of 89 years. He was born in New York, Oct. 20, 1806. His wife, to whom he had been married more than sixty years, died only five days earlier.

Mr. Lawrence was one of the most careful and prolific of American ornithologists. The list of his published writings contains 121 titles, the earliest of which appeared in 1844, the latest in 1891. The period of his productive activity thus covered nearly half a century. He was an active contemporary of all American ornithologists from Audubon and Nuttall to the younger writers of the present day. Baird, Cassin and Lawrence' are classic names in ornithology -names associated in joint authorship in Baird's great work on the birds of North America, published in 1858. For nearly fifty years Baird and Lawrence, then the foremost authorities on American birds, were warm personal friends, and on more than one occasion accomplished, by hearty coöperation, what neither could have done alone. It should not be forgotten that their arduous labors paved the way for the refinement of detail that characterizes the bird work of to-day.

Baird busied himself chiefly with the birds of the United States, Lawrence chiefly with those of tropical America. Lawrence described more than 300 new species from the West Indies, Mexico, Central and South America. One genus and twenty species were named in his honor-tokens of respect and esteem-by American and European naturalists.

Baird and Lawrence lived under widely different conditions. Baird led an active official life, burdened with the cares and responsibilities of three great institutions, two of which, the National Museum and Fish Commission, were his own creation; he was constantly overworked and died prematurely at the age of sixty-five years. Lawrence led a quiet, retiring life, far away from the public eye, and died at the ripe age of fourscore years and nine. Still, the

* The Published Writings of George Newbold Lawrence, by L. S. Foster. Bull. U. S. National Museum,

two had many traits in common; both were plain and unassuming, kind and thoughtful in their family relations, and ever ready to extend a helping hand to those, however young, whose tastes led them to the study of birds. In looking back over the twentyfive years that have passed since I first enjoyed their acquaintance, my mind constantly recurs to the kindly words of encouragement and advice that shaped my early course as a naturalist, and the friendships that followed will always live among my most cherished memories.

C. HART MERRIAM.

SCIENTIFIC LITERATURE.

A Treatise on Hydrostatics. By ALFRED GEORGE GREENHILL, Professor of Mathematics in the Artillery College, Woolwich. Macmillan & Co., London and New York. 16mo, pp. viii+536.

The science of hydrostatics, originating with Archimedes, is now more than twenty centuries old. It is, in many respects, one of the most perfect and satisfactory of the sciences. This fact, however, arises from the simplicity of the phenomena with which hydrostatics has to deal rather than from anything like continuity of progress during its lengthy history. Indeed, as regards purely hydrostatical principles, we are not very greatly in advance of Archimedes. Our superiority over him is due, first, to an immensely enlarged capacity, through the developments of mathematics, for the application of those principles; and, secondly, to the exploration of the much larger and more interesting domain of hydrodynamics, of which, in fact, hydrostatics is only a special case.

The work of Professor Greenhill treats hydrostatics from the modern point of view. He does not hesitate to cross the border for an excursion into hydrokinetics whenever desirable or essential, although some might

infer from the title of the book that such excursions are avoided. The scope and character of the work may be best inferred from the following paragraphs of the preface:

"The aim of the present Treatise on Hydrostatics is to develop the subject from the outset by means of illustrations of existing problems, chosen in general on as large a scale as possible, and carried out to their numerical results; in this way it is hoped that the student will acquire a real working knowledge of the subject, while at the same time the book will prove useful to the practical engineer."

"In accordance with modern ideas of mathematical instruction, a free use is made of the symbols and operations of the Calculus, where the treatment requires it, although an alternative demonstration by elementary methods is occasionally submitted; because, as has well been said, "it is easier to learn the differential calculus than to follow a demonstration which attempts to avoid its use."

Too much stress cannot be laid on this remark with regard to the rôle of the calculus in applied science. We are coming now, after two centuries, to realize clearly that the use of the calculus has become general in all higher investigations, not because the pure mathematicians have so desired, but because the phenomena of nature demand for their interpretation such an instrument of research.

The book is a mine of interesting and useful information, and must become one of the standards for students, teachers and engineers. The principles are illustrated by a wide variety of good examples, many of which are drawn from practical applications. Special attention is given to the problems of flotation and stability of ships, and to problems arising in naval architecture. The theory of the various hydrostatic instruments, including the hydrometer, the barometer and the gas thermometer, is

worked out quite fully. A chapter is devoted to pneumatics, and another to pneumatic machines. There are also chapters on capillarity, hydraulics, the general equations of equilibrium, and on the mechanical theory of heat. In short, the work is a very comprehensive one. Few books contain more information per page, and few abound to such an extent in historical references.

The exposition of the author is in general clear and logical, though occasionally an important principle is announced without due warning. Thus, Bernoulli's theorem appears without demonstration on p. 467 in the chapter on hydraulics. It would have been more in accord with the admirable spirit of the book, we think, if the author had given in that chapter the general equations of fluid motion, and thence deduced Bernoulli's theorem, even if this enlargement had required a change in the title of the work.

Some obscurity arises here and there from the author's habit of condensation. Thus, on p. 458 we read, "so that the attraction of pure gravitation on a plummet weighing W g is WG dynes, where G denotes the acceleration of gravity." Of course, the expert would quickly see that Wg means W grammes, but the average engineer will not commend such economy.

The book has a good, but not quite good enough, index. For example, the unusual words barad and spoud are occasionally used by the author. Their meaning is plain from the context, in most cases, to the specialist, but the general reader would not get any light on these terms from the index; for it does not contain the word spoud, while it refers for barad to a page on which this word does not occur.

These faults, however, are small ones, and such, moreover, as are well-nigh inseparable from the first edition of a book so full of sound knowledge as this one.

R. S. WOODWARD.