404 H. H. NEWMAN 



somewhat more abnormal head and eyes and a predominance of 

 the paternal tendency in chromatophores, mackerel greens being 

 much more numerous than Fundulus reds. 



Class F. A medium-sized group of eighteen individuals that 

 are decidedly more subnormal, especially in apical parts, than 

 class E. These individually show no differentiation of eyes, 

 no otic vesicles, and no heart. The posterior parts of the body 

 varj^ from quite long slender posterior prolongations to shorter 

 and broader processes. In all of these the mackerel types 

 of chromatophores are at least as prominent as the Fundulus 

 types. In many cases the former greatly predominate over the 

 latter and in four individuals one has to make a careful search 

 to find any chromatophores that are definitely of the maternal 

 type. These last individuals are as pure paternal as some of the 

 hybrid embryos that Loeb described as pure maternal. 



Class G. A group of sixteen individuals that are little more 

 than amorphous masses of living cells. Only the transparency of 

 the masses and the expanded condition of the chromatophores 

 show that these amorphous embryos are still living. While 

 the majority of these show distinct biparental inheritance of 

 chromatophore types, two are absolutely pure paternal. 



Class H. Twelve embryos that had begun cleavage have died 

 and have been removed from time to time. In no case had they 

 developed beyond early gastrulation. 



Excluding the dead embryos, there are seven groups that 

 usually overlap somewhat so that a continuous series might 

 readily be made. The arbitrary classification, however, is a 

 necessary part of the experimental method, for each class was 

 put into a separate vessel and followed throughout the life of the 

 embryos. For about a week after the segregation into classes 

 was made only occasional studies of selected individuals were 

 undertaken and a good many drawings were made for record. 

 On July 1, however, nineteen days after fertilization, the seven 

 classes were subjected to a second complete census with the 

 following findings : 



Class A. One individual had hatched on the previous day 

 after eighteen days' development; four or five days late as com- 



