THE AMERICAN NATURALIST. [Vol. XXXV I II. 



can be made, however, for this 42-cell stage, since but a single 

 specimen was available for study, but it may be seen from the 

 above sketch of this and the preceding stages that the eggs of 

 Desmognathus will furnish a very interesting and convenient 

 object for the study of cleavage problems in Amphibia. 



Sketches of two later cleavage stages are given in Fig. 2 and 

 may be seen to consist of a greater and greater subdivision of 

 both micro- and macromeres, apparently without special regu- 

 larity. The last figure (June 24) represents the lower pole of 

 the oldest stage I succeeded in obtaining. 



The' above obser\'ations. which establish the fact that the 



©C €1 



[)gnathus are holoblastic, will f 

 statement to the opposite effect given in my previous article on 

 the sul)ject, and while it is always unpleasant to be proven in a 

 niis-statement, it is more satisfactory to the investigator to be 

 able to furnish his own j^roof than to leave it to others. At the 

 time ot writing niy ])revi()us article, I possessed no embryos 

 younger than those represented in the upper row of Fig. 3, and 

 It then appeared to nie impossible that such a relation between 

 embrvo and yolk as the one shown here could result from an egg 

 of the holoblastic type. A slightly older stage, that represented 

 ot hig. 3. seemed still more convincing in 

 this time a set of branching blood vessels 



this resnec 



