462 NATURAL SCIENCE. June. 



food. Fig. 7 represents an imaginary stage in the development of a 

 form which has lost food-yolk, and which, in consequence, has a large 

 hollow segmentation cavity into which follicle-cells are migrating 

 from the layer of follicle-cells surrounding the embryo. In Fig. 8 is 

 represented the invaginating stage of such an embryo. The follicle- 

 cells are, of course, inside the segmentation cavity, but they also 

 follow the invagination, and fill the hollow that usually forms the 

 digestive tract. Now in Fig. g is represented a similar stage, but in 

 a case where the follicle-cells, instead of migrating irregularly, 

 arrange themselves as a regular epithelium. In it is to be seen the 

 ordinary invaginate gastrula, but the cells derived from the follicle 

 arrange themselves as a regular cast of this. In subsequent develop- 

 ment it is easy to see that the follicular epithelium might follow all 

 the developments and changes of the real embryonic cells, so that 

 each fold of tissue would be formed not only of embryonic cells, but 

 of follicle-cells closely adherent. Dr. Brooks supposes that Salpa had 

 such a history, and that, when this condition of things was reached, 

 the development of the embryonic cells began to lag behind the 

 development of the follicle-cells, which originally merely followed the 



course of the cells they came to 

 /0 ^f|-/-7 777?-r>^ anticipate. In Fig. lo is shown the 



/S^h^'Jyf^^r'^^^<^y^>^ case of an embryo in which the 



l^\i--' ' ■'"'^^y^'^^'^ld^^^ invagination has taken place in 



-^^ j^' '/®}-(®)"./'s^ )^ follicle-cells, but where the real 

 Bf ^f®)^^'''^'^'^^ ^ embryonic cells are so delayed that 

 ^ \^>^-'C®^'/Q\\\?y J^ they appear merely as scattered 

 i^k X:/.//rM l^''~2^>^><w^''^ segmentation spheres, and this last 



XV>s,^^^ ''' i^i ' ^ °l -''''^O^S^ figure is merely a diagrammatic 



x^^>r- ) N ' M I \ '\S^^^^^ representation of what actually 



^^^^^^'^■^-^-^^^ occurs in Salpa, as may be seen by 



• ^'u- ^°-Diag''^"^ °f .a" «"?bryo comparison with Fig. 5. 

 in which the segmentation spheres ^t^, , 1 j j • 



delay development. There has already appeared m 



Natural Science, vol. iii., p. 223, a 

 brief account of the interesting views to which the development of 

 Salpa has led Dr. Brooks. As that account was based on three 

 chapters of this monograph which appeared separately, it is 

 necessary now only to remind readers that Dr. Brooks has come 

 to the interesting conclusion that a simple pelagic form, of which 

 Appendicnlavia is the nearest living representative, is the common 

 ancestor of the Ascidians and the Chordates ; and that consequently 

 there can be only among such organs as are found in simple transparent, 

 surface-living creatures any homologies between vertebrates and 

 invertebrates; arguments linking annelids, Crustacea, and so forthwith 

 vertebrates, because of similarities in segmentation, body-cavity, etc., 

 are dismissed as empty shadows of morphological dreams. 



In this short attempt to bring to the notice ot the general 

 scientific public some of the more interesting results of Dr. Brooks' 

 work, it has been impossible to touch upon many of the more 

 complicated and technical matters which all specialists will have to 

 study with attention in this great contribution to scientific literature. 



Ferns. 

 The Study of the Biology of Ferns by the Collodion Method, for advanced 

 and collegiate students. By G. F. Atkinson, Ph.B. 8vo. Pp. xii, and 134. 

 New York : Macmillan & Co., 1894. Price Ss. 6d. nett. 



The author, who is associate professor of cryptogamic botany in 

 Cornell University, Ithaca, N.Y., was led to the preparation of this 



