The Morpholog}^ of Cosmobia 437 



another early conclusion^ namely, that the "lens is formed from 

 ectoderm different in position from that of the normal lens-forming 

 region" for in each of the specimens figured each lens component is 

 perfectly normal with reference to the associated elements of the optic 

 cup, and it is more likely that the ectoderm from which these were 

 formed was that which belonged in front of the eye-anlage from the 

 start, and that the parts missing, which normally appear between the 

 eyes, consist as well of integument as of the other tissues. In the mam- 

 malian Cyclops monsters which I have described in the body of the 

 text it will be remembered that this interocular integument is rolled 

 in as the lining of a median canal, while the other parts appear either 

 in the walls of this canal or above the componental eyeball. It would 

 then seem that the assumption that the ectoderm which here develops 

 into the lens components, was originally inter- or prse-ocular. rests upon 

 no definite basis. 



The dependence of this theory of a shifting lens-anlage upon that of 

 a migrating eye-ball was also seen by Dr. Stockabd during the progress 

 of his work and in a recent number of Science (Oct. 2, 1908) he corrects 

 this early position by means of the following results. It seems that in 

 a large number of the monsters produced the defect is not cyclopy but 

 the reduction or loss of one of the normal lateral eyes, the other being 

 normal in all respects. In many of the monsters of this type a lens 

 develops on the eyeless side, showing that a lens may develop quite 

 independently of the stimulus presumably afforded by the presence of an 

 optic cup; /. c. the lens is "self-differentiating.'' This standpoint is 

 quite the opposite of his earlier one, which claims that a lens develops 

 wherever a migrating optic cup comes to rest, but is a necessary conse- 

 quence of a change of position with respect to the fusion question. 



Aside from the theoretical side, however, there remains the fact 

 that the author, in a series of brilliant experiments, has been able to 

 produce experimentally what appear to be genuine Cyclopean monsters, 

 and that, since his method is a chemical rather than a mechanical one. 

 and therefore more subtle in its action upon minute parts, his results 

 have been similarly more perfect than those usually employed for the 

 production of double monsters. Since, however, the chemical medium 

 seems to have been applied not later than the four-celled stage, at which 

 time the blastomeres may not have become differentiated, its action 

 may have affected the germinal mechanism itself rather than the 

 developing embryo, and thus farther investigation will be necessary 

 before we can point out the true factor in the development of defective 



