The Embryology of Heterodontus japonicus 727 



These furrows are \nown to occur in no other Selachian. — Compare, however, the segmeiT 

 tation stages of Chimaera [Dean, 1906] , and take into account the paleontological history of 

 Heterodont sharks. In a word it is precisely since they do occur in Heterodontus that these 

 furrows may well be homologous with cleavage lines. 



The furrows have not been traced bac\ into the earliest segmentation stages. — A gap in the 

 evidence, truly, but by no means a fatal one. In earliest stages examined a continuity has 

 been shown between the inter-blastomeral spaces in the germ and the circum-germinal furrows. 



The furrows may have no relation to the nuclei. — We note however, that the furrows do 

 not occur in the egg about the time of fertiliziation: i.e. prior to segmentation. It has been 

 demonstrated that nuclei are abundant in the region beyond the germinal area, to a distance 

 of about 10 degrees on all sides. We have some reason to infer that they extend further peri- 

 pherally since the neighboring circum-germinal yolk is similar in character to that in the 

 region where they occur. Unhappily, however, owing to technical difficulties, the outer re 

 gion of the yolk has not been sectioned. But it does not follow that, because nuclei in this 

 region have not been demonstrated, the furrows in question cannot be concerned with 

 cleavage. For such an objection would apply equally well to the case of such eggs as those of 

 Jiecturus where nuclei have not been demonstrated in the vegetal region, yet where one does 

 not question the homology of the furrows with cleavage lines. 



The furrows may he due to the action of merocytes, which are \nown in Pristiurus and 

 Torpedo to form blastomere-li\e structures. — Even in this event the furrows must be classified 

 broadly, I think, as within the category of cleavage lines, and hence as an expression of a holo- 

 blastic condition. For if an egg subdivides, when deprived of its nucleus and later provided 

 with a sperm nucleus, does not this division come under the general head of cleavage? There 

 is, however, no evidence that furrows of so distinct a type have ever been produced in 

 a meroblastic egg by merocytes. There is on the contrary evidence for assuming that mere 

 cytic division in Cestracion [Heterodontus] would be less evident than in more modern types of 

 sharks. For all will agree that polyspermy, in vertebrates at least, is a secondary character and 

 less apt, therefore, to have been prominent in the oldest sharks, like Cestracion. Indeed, we 

 already know, thanks to Ruckert's studies (1899) that the migration of merocytes into the 

 yolk is less marked in Pristiurus (an older form) than in Torpedo (a later and derived form). 



We conclude, accordingly, that the weight of the evidence is unquestionably in favor of 

 regarding the furrows in the early Cestraciont egg as the homologues of cleavage lines. 



Among Dean's records I find several photographs of eggs exhibiting the alleged 

 "holoblastic" cleavage furrows. These photographs are in part identical with those 

 published in Dean's article (1901) on cleavage. Some of the photographs show the 

 furrows quite as clearly as they are portrayed in Dean's drawings. But it is not likely 

 that these drawings were made from photographs. Inserted in Dean's notebook there are 

 many carefully executed drawings of these "cleavage" stages, annotated in Dean's 

 handwriting. Some correspond to the drawings already published; but few, if any, 

 correspond to the photographs. 



In some of the drawings found in Dean's notebook the region of the upper pole of 

 the egg, which Dean sometimes calls the animal pole, has the appearance of a well-defined 

 large blastoderm or region of micromeres, from which nearly parallel furrows radiate like 

 meridians down over the equatorial region (as in Figures 1 to 6, plate I). Thus the egg has 



