The Morphology of Cosmobia 427 



the physiological processes hindered. Such deformity, which may 

 be considered as secondary, is very much more frequent in the 

 abnormal than in the normal Cosmobia since both the surroundings 

 during development and the anatomical mechanism are fitted for the 

 usual type. It may even happen that certain abnormal types, which 

 are theoretically possible, cannot come to full development, being 

 always and inevitably stopped at a definite period by some mechani- 

 cal difficulty which it is impossible to overcome. These difficulties 

 are different in different animals, since the anatomical structure 

 and the mode of development is different, and this may account for 

 the frequency of certain types of monsters and the non-appearance 

 of others in certain animals or animal groups. 



4. As in normal, so in abnormal Cosmobia, the ultimate cause of the 

 development of the organism and its architectural details lies in 

 the germ, and is probably determined by the time of the first 

 cleavage. There is thus neither a "fusion" of parts already formed 

 nor a gradual development from the normal towards the abnormal 

 during embryonic life, but the parts appear doubled or reduced 

 from their first appearance and their development is controlled in 

 the same way as are the bilateral structure and other architectural 

 characteristics of normal beings. 



5. If this latter be true, and it is the most hypothetical part of these 

 conclusions, it naturally follows that true excessive or defective 

 Cosmobia can be produced experimentally only through a cause 

 which is applied early enough to lead to the formation of a germ 

 that differs as much from the normal type of germ as the resulting 

 organism differs from the normal adult. This would lead us to 

 question the value as true Cosmobia of such experimentally produced 

 organisms as the double embryos that result from the partial separa- 

 tion of the early blastomeres ; and as a matter of fact the results of 

 such experiments, when critically examined, are hardly symmetrical 

 enough to be called Cosmobia, and although extremely suggestive, 

 give the impression of monsters produced, not by the same cause 

 as that employed by Nature, but by a similar one, which secondarily 

 duplicates that part of the mechanism of control which in the germ 

 of a true Diplopage is double from the start. In the more recent 

 experiments with Teleost eggs (Stockard's) in which defective 

 monsters have been produced by the introduction into the water of 



