OF ARTS AND SCIENCES. 53 



features of this supposed horizontal cleavage ; and Hoffmann will 

 probably agree with us that the circumstantial evidence ought to be 

 more complete, before we concede such a fundamental difference in 

 development between closely allied fishes. 



If, however, it be claimed that Hoflfmann's figures furnish conclu- 

 sive Evidence of his view, we shall have to admit that the first ap- 

 pearances favor this claim ; but a somewhat closer examination of the 

 text and the figures leaves a very different impression. A fair pre- 

 sentation of the question at issue compels us to call attention to Hoflf- 

 mann's Plate IV. Figs 1-4, and the explanation of the same as given 

 in the Tafel-Erkliirung (p. 165) and in the text (p. 106). As it does 

 not comport with the purpose and limits of this paper to give a com- 

 plete historical sketch of his observations on the origin and develop- 

 ment of the periblastic layer, we have selected for examination the 

 pl.Hte which most fairly represents the grounds of his view. 



Hoffmann states, in a very plain and direct manner, that Fig. 1 

 (Plate IV.) represents a 2-celI stage of his archiblast ; we think this is 

 what we have called the 4-cell stage, but cannot affirm this positively. 

 It is also explicitly asserted that Fig. 2 represents a 4-cell stage 

 (of the archiblast). Now this is assuredly an error. There are only 

 two stages in the whole development which could give a view approxi- 

 mating that seen in this figure; namely, the 8-cell and the 16-cell 

 stage. As this figure is only aboxit five minutes later than Fig. 1, 

 according to the explanation on page 165, it is quite impossible that 

 Fig. 1 should represent anything earlier than our 4-ceIl stage. The 

 next statement is certainly astounding ; for it declares that Fig. 3 

 represents the 8-cell stage (we are speaking of the "archiblast"). 

 We can assert with well-founded assurance that this figure cannot be 

 said to represent any stage earlier than the 32-cell stage. The climax 

 is reached in the twice-repeated statement (pp. 106, 165) that Fig. 4 

 represents a stage in which the "archiblast" is composed of 16 cells. 

 Comment is unnecessary. Allowing that, as inadvertencies, these 

 statements do not completely invalidate the figures, it may still be 

 fairly claimed that they raise grave doubts as to the accuracy of Hoff- 

 mann's interpretation ; and that they furnish us with a good reason 

 for setting aside this interpretation, provided we can replace it with 

 one that is more satisfactory. 



We regard Fig. 1 as a 4-cell stage seen somewhat obliquely, so 

 that two nuclei appear below the other two, as if they were in a 

 subjacent stratum of protoplasm. The figure is illusory, and gives no 

 idea of the relation of the blastodisc and the periblast at this stage. 



