LAURABEE. — THE OPTIC CIIIASMA OF TELEOSTS. 227 



Group I includes the double-headed specimens which were divided 

 only anterior to the dorsal fin. 



Group II includes double-headed specimens which were divided 

 posterior to the dorsal fin, but had a portion of the vertebral column 

 in common. 



Group III includes specimens in which the two bodies were united 

 throughout a portion or the whole of the tail region, but were else- 

 where distinct. 



Group IV includes specimens with two young entirely distinct and 

 on opposite sides of the yolk sac. 



The bearing which these dicephalous specimens have upon the 

 subject under discussion can be shown by briefly considering some 

 of the theories advanced and some of the results obtained by those 

 who have made a study of this condition. Among other causes of 

 dicephaly, polyspermy has been advanced by several writers as a 

 probable cause. If this occurred, it is evident that the heredity of 

 these monstrosities would be more complicated than that of normal 

 individuals. 



Schmitt (: 02), in a paper on the gastrulation of double trout em- 

 bryos, stated that the germinal area of the egg which produced a 

 double embryo was no larger, and contained no more material, than 

 the one which gave rise to a normal trout. In the formation of the 

 gastrula two invaginations occurred simultaneously in different parts 

 of the blastoderm. This observation indicates that polyspermy had not 

 occurred ; for if it had occurred, there would have been an increase in 

 the material contained in the germinal area. This change would also 

 occur if the egg were fertilized by a double-headed sperm, one of the 

 tentative causes for duplicity advanced by some writers. 



Hans Spemann (:01, :01*), in experimenting with the eggs of the 

 salamander Triton, was able to produce dicephalous individuals, or even 

 two distinct individuals, by means of a ligature of hair tied across the 

 blastoderm. The extent of the duplicity depended on the tightness 

 of the ligature. A similar result has been obtained by Loeb and 

 others, although by different methods. 



From these various observations it is evident that an increase in 

 the amount of sperm, either by polyspermy or otherwise, is not, at 

 least commonly, the cause of dicephaly. Accordingly in two-headed 

 trout, the hereditary characters of the divided parts should be the 

 same. It would follow, then, that if the dimorphic condition is 

 hereditary, the condition of the crossing should be the same in each 

 head. By referring to Table VIII, it will be seen that in more than 

 one half of the double-headed specimens dissected the crossings in the 



