ous localization of germ-plasm in plants. Weismann himself admits that the germplasm must be distributed in minute fraction' in all 'somatic nuclei' of the begonia leaf, because that leaf is capable of giving rise to new plants, by means of cuttings, and all the plants may produce good flowers, which, if they are sexual at all, are so only by virtue of containing some of this elusive germ-plasm. There is no other way for these plants to get their germ-plasm, except from the somatic leaf from which they came. It would seem that this admission undermines the whole theory of the localization of the germ-plasm in plants, for one exception in the hypothesis must argue that there are others. But not so! There are no insurmountable difficulties before the Weismannians. It is the begonia which is the exception, for it is abnormal for plants to propagate by any such means! The an

swer which has been made to this statement is that very many plants are propagated asexually by horticulturists, and that all plants can probably be so propagated if there were any occasion for the effort. This answer is true; but the philosophical answer is that every phyton is an autonomy, and that the mere accident of its growing on the plant, in the soil, or in a bottle of water, is wholly aside from the point, for wherever it grows it lives at first a sexless life, it has an individuality, competes with its fellows, varies to suit its needs, and is capable, finally, of developing sex.

Another fundamental tenet of Weismannism is the continuity of the germplasm, the passing down from generation to generation of a part or direct offspring of the original germ-plasm. Now, if there is any continuity in plants, this ancestral germ-plasm must be inextricably diffused in the soma-plasm, as I have said, for every part or phyton of these plants, even to the roots and parts of the leaves, is able to produce sexual parts or germ-plasm. And if

this germ-plasm is inextinguishably associated with every cell of the plant body, why does it not receive and transmit all incident impressions upon the plant? Why should acquired characters impress themselves upon the soma-plasm and not upon the germ-plasm when this latter element is contained in the very nuclei, as Weismann admits, of somatic cells? If the theory of the continuity of the germ-plasm is true for plants, then acquired characters must be transmitted! The only escape from this position is an arbitrary assumption that one plasm is impressionable and that the other is not; and, now, that we can no longer relegate the germ-plasm to imaginary deepseated germ-cells, such an assumption is too bold, I think, to be suggested.

The entire Weismannian hypothesis is built upon the assumption that all permanent or progressive variation is the result of sexual union; but I have shown that there is much progressive variation in the vegetable kingdom which is purely asexual, and, for all we know, this type of modification may proceed indefinitely. There is no doubt of the facts; and the only answer to them which I can conceive the Weismannian to make is that these progressive variations arise because of the latent influence of ancestral sexual unions. In reply to this I should ask for proofs. Hosts of fungi have no sex. I am not convinced but that there may be strains or types of some species of filamentous algæ and other plants in which there has never been sexual union, even from the beginning. And I should bring, in rebuttal, also, the result of direct observation and experiment to show that given hereditable asexual variations are often the direct result of climate, soil or other impinging conditions. As a matter of fact, we know that acquired characters may be hereditary in plants; if the facts do not agree with the hypothesis, so much the worse for the hypothesis. Unfortunately,

the hypothesis is too apt to be capable of endless contractions and modifications to meet individual cases. I sometimes think that we are substituting for the philosophy of observation a philosophy of definitions. I have, therefore, attempted to show : 1. That the plant is not a simple autonomy in the sense in which the animal is.

2. That its parts are virtually independent in respect to (a) propagation (equally either when detached or still persisting upon the parent plant), (b) struggle for existence amongst themselves, (c) variation, (d) transmission of their characters, either by means of seeds or buds.

3. That there is no essential difference between bud-varieties and seed-varieties, apart from the mere fact of their unlike derivation; and the causes of variation in the one case are the same as those in the other.

4. That all these parts are at first sexless, but finally may or may not develop sex.

5. That much of the evolution of the vegetable kingdom is accomplished by wholly sexless means.

There is, then, a fundamental unlikeness in the ultimate evolution of animals and plants. A plant, as we ordinarily know it, is a colony of potential individuals, each one of which, save the very first, is derived from an asexual parent, yet each one may, and usually does, develop sex. Each individual is capable also of receiving a distinct or peculiar influence of the environment and struggle for existence, and is capable, therefore, of independent permanent modification. It is not possible, therefore, that there is any localization or continuity of a germplasm in the sense in which these conceptions are applied to animals; nor is it possible for the plant as a whole to make a simple functional adaptation to environment. If there is a continuity of germplasm in plants this element must of necessity be intimately associated with every par

CURRENT NOTES ON PHYSIOGRAPHY (III.). WOODWARD'S SMITHSONIAN GEOGRAPHICAL TABLES.

'THE average geographer,' to whose needs Professor Woodward has attempted to suit the recent volume of Geographical Tables issued by the the Smithsonian Institution, should certainly feel highly complimented by this tribute to his quality. The volume contains, among many other matters, tables of coördinates for the projection of polyconic maps, lengths of a degree on parallels and meridians at different latitudes, areas of latitude-and-longitude, quadrilaterals of different dimensions and at different latitudes, adopted dimensions of the earth's spheroid, value of gravity at the earth's surface, and salient facts of physical geodesy. The latter heading includes the area of the earth, of oceans and continents, and the average heights of continents and depths of oceans, taken from Helmert's Geodäsie. For areas the continents are given 51,886,000, and the oceans 145,054,000 square miles. The mean depth of the oceans is placed at 3,440 meters. The mean heights of the continents are given as follows: The earlier results of Humboldt's, still often quoted, and the later ones of Penck (Morphologie der Erdoberfläche, 1894) being added for comparison.

The increase in the values of the latter measures is probably an approach to the truth, for early explorations frequently gave too much emphasis to narrow mountain ranges, and too little to broad plateaus.

A. AGASSIZ ON THE BAHAMAS.

A RECONNOISSANCE of the Bahamas and of the elevated reefs of Cuba, made by A. Agassiz in the winter of 1893, affords material for a Bulletin of 200 pages with 47 plates and many figures in the text, lately issued by the Museum of Comparative Zoology at Harvard College. The author is emphatic in rejecting the sufficiency of the Darwin-Dana theory of submergence in explaining the features of great limestone banks. The Bahamas consist of low hills of æolian limestone, "formed during a period of rest, during which the great beach of the then existing reef constantly supplied fresh material to be changed by the surf and the winds into sand for the heaping up of sand dunes. They could not be formed in a district of subsidence unless the subsidence was slower than the rate of growth of the corals, which is not the case in the Bahamas, as the reefs of to-day, even when they come to the surface, are not the sources from which the material for the great dunes of the Bahama Islands is derived" (p. 184, 185). At present the dunes are disappearing before the action of the sea. The conclusion of the reconnoissance seems to be that the great limestone banks are chiefly formed as 'marine limestones,' accumulating at great depths by accretion;' and that in the West Indies" wherever coral reefs occur, and of

whatever shape, they form only a comparatively thin growth upon the underlying base" (p. 177). The text, with its figures, supplemented by maps and plates, gives an excellent idea of the geographical features of the region and of their evolution.

SPENCER'S RECONSTRUCTION OF THE ANTIL

LEAN CONTINENT.

PROF. MARCEL BERTRAND, of the École des Mines and the Geological Survey of France, has published an account of certain faint deformations of northwestern France, in which he interprets the inequalities in the floor of the English channel as the result of faint anticlinal and synclinal movements (Bull. Soc. Géol. France, xx., 1892, 118); thus implying that neither erosion nor deposition has been of significant measure in shaping the channel floor. Prof. J. W. Spencer takes almost the other extreme, and interprets certain inequalities of the ocean floor of the Antillean region, even to depths of twelve or fifteen thousand feet, as the results of river erosion during a not remote time when the entire region is supposed to have had a much greater altitude than at present (Bull. Geol. Soc. Amer., vi., 1895, 103-140); thus implying that no other processes than river erosion can account for the inequalities that he has traced. It must be concluded from these contrasted arguments that the forms of the sea floor are not yet so well understood as those of the land; because the facts are much less accurately known under than over sea level, because only form and not structure can be determined by soundings, and because the forms of the sea floor have received relatively little study. Where two specialists reach conclusions so unlike, it is difficult for others to choose between them; and for the present there will probably be some hesitation in adopting the teachings of the one or the other. With much interest aroused in the facts brought forward, and

with all willingness to look on the continents as unstable, it is difficult to believe that they have suffered changes so great as Spencer announces, not only in the uplift of the Antillean region, but in the deep depression of the axis of Central America, and in the denudation of the (inferred) great banks or continental shelf along the Windward Islands. The strongest proof will be demanded before vertical movements of two miles and a half can be accepted; and we fear that most readers will take refuge in a verdict of 'not proved.'

HISTORY OF THE ST. JOHN RIVER, NEW
BRUNSWICK.

AN article on the 'Outlets of the St. John river,' by G. F. Matthew (Bull. Nat. Hist. Soc., New Brunswick, xii., 1894, 43-62), concludes that this river has been built up by contributions from three other systems, whose lower parts are now to be seen in the Restigouche, Miramichi and Petitcodiac. The evidence of this conclusion is derived from the geological structure of the country, beginning as far back as the Huronian time; the three rivers whose upper basins now belong to the St. John having been defined as basins of deposition in more or less remote geological periods. Thus the St. John river has attained its present magnitude by the breaking of mountain or hill barriers which once separated its three river systems, and is not a simple valley of continuous growth like the Mississippi (p. 55). The difficulty of accepting Dr. Matthews' conclusion as the only solution of the history of the St. John does not lie in any objection to the geological history of the region and its several basins of deposition, as far as stated, but in the omission of sufficient consideration of what has happened in the region since it became a land area. It has long been subject to subaërial erosion. During this time it has in all probability been variously warped and

otherwise moved with respect to its baselevel. Its rocks are of diverse resistance, and hence there may have been repeated opportunities for diversion and rearrangement of river courses during the long life of the region as a land area. While admitting that several geological basins of great antiquity are now drained by a single river, it does not necessarily follow that this river is an immediate descendant of the rivers which at one time or another drained the separate basins. The actual St. John river may once have been larger than now; its neighbors may have gained drainage area from it instead of losing drainage area to it; but these possibilities are not considered.

THE ORIGIN OF THE MISSISSIPPI.

THE reference to the Mississippi in the previous paragraph brings up an oft-encountered implication of simple history in the development of this great river, against which there is much evidence. A similar implication is found in a recent State Survey Report, where it is stated that, as a result of continental evolution at the close of the Carboniferous period, the drainage of the Ohio region was turned southward "into the great Mississippian bay, which then washed the shores of the new-born continent as far north as the mouth of the Ohio river" (Geol. Coastal plain of Alabama, 1894, 11). It is found again in the 'Story of the Mississippi-Missouri,' where the Mississippi at the close of the Appalachian revolution is described as heading somewhere in the Minnesota-Wisconsin region, and flowing southward to its mouth somewhere near the present city of St. Louis, whence a deep gulf extended southward to the present Gulf of Mexico (Amer. Geol. iii., 1889, 368). While the southwardflowing streams of the Wisconsin-Minnesota highlands are probably of ancient origin, the southward course of the Mississippi between Tennessee and Arkansas seems to

have been initiated not at the close of the Appalachian revolution, but long afterwards in Cretaceous time. The Appalachian revolution formed the mountains of Arkansas, as well as those of the Alleghany belt. The similarity of structure is so great that a trans-Mississippian extension of Appalachian growth may be reasonably assumed, as has been pointed out by Winslow (Bull. G. S. A., ii., 1891, 231). The existence of a bay, from the Gulf of Mexico northward towards St. Louis, is very improbable as a result of the Appalachian revolution; an east and west constructional mountain belt is a more likely product; and not until this mountain belt was well denuded to a peneplain did a later deformation depress it transversely, admitting the Cretaceous waters northward across it, and thus first forming the Mississippi embayment. Probably in part at the same time, and to a greater extent in later time, the denuded peneplains to the east and west were raised towards their present upland altitude, and as a result of this elevation the existing valleys and lowlands were opened in them during some part of Tertiary time. With these later elevations we may associate the uplift of the filled embayment and the southward growth of the Mississippi as a river. This view of the origin of the Mississippi embayment and of the date of the southward discharge of Mississippi drainage was first published by L. G. Westgate (Amer. Geol. xi., 1893, 251), as a result of conference with L. S. Griswold, who had then recently completed his investigation of the novaculite region of Arkansas.

THE CHUNNENUGGA RIDGE AND THE BLACK PRAIRIES OF ALABAMA.

Ir is, perhaps, too much to expect that the origin of the physiographic features of a region should always receive due attention in a geological report along with the origin of its strata; yet there is no other place so

.

appropriate for the official publication of physiographical discussions. It therefore occasions regret to find so little account of the origin and meaning of the Chunnenugga ridge and the Black prairies of Alabama in the elaborate report on the Geology of the Coastal Plain lately published by the Survey of that State. of that State. "The Chunnenugga ridge is made in great part by alterations of hard limestone ledges and bands of indurated sands of the Ripley. It overlooks the low trough of the black prairies of the Rotten limestone towards the north with somewhat precipitous slopes in that direction, while its descent towards the south is much more gentle " (p. 356). It is manifest that the ridge with its inland-facing escarpment and the denuded inner lowland are typical features of a certain stage in the denudation of a coastal plain that consists of more and less resistant strata; the drainage of the lowland being chiefly gathered by subsequent streams that have been developed along the strike of the beds, and discharged by consequent streams which maintain transverse valleys through the enclosing ridge or upland. This general relation of form and drainage is so often repeated on coastal plains that its occurrence in Alabama deserves mention as a local example of a general physiographie feature; just as the Cretaceous strata on which it is developed deserve mention as local examples of a widespread geological formation.

HARVARD UNIVERSITY.

W. M. DAVIS.

THE NEW YORK MEETING OF THE ASSOCIATION OF AMERICAN ANATOMISTS. THE Seventh Annual Session of the American Anatomists was held in the Medical Department of Columbia College, 437 West 59th Street, New York City, December 28 and 29, 1894.

The Association was called to order Friday, December 28th, by the President, Dr.