260 PHYSIOLOGICAL GENETICS 



A real pattern process is involved in the goldfish (Carassins 

 auratus). Here black and yellow chromatophores arc present, 



and the mottled patterns arc caused by a secondary destruction 

 of pigment cells in definite areas (Berndt, 1925; Fukui, 1927, 

 1930). The resulting type depends upon the time of beginning 

 of destruction, place and amount of destruction, and whether or 

 not also yellow chromatophores are finally destroyed. Thus a 

 seriation brown self — variegated black and orange — orange — 

 variegated orange and white — white may be obtained. The 

 pattern itself is quite irregular and is complicated by the presence 

 of a pattern of scales with and without (transparent-type) 

 guanin crystals. There is, according to Fukui, a relation between 

 the pattern of chromatophore destruction and the structure 

 of the skin, and the destruction of pigment is combined with 

 phagocytic action. Goodrich and Hansen (1931), who made a 

 quantitative study of these facts, point out that most of them 

 may be explained by a system of reactions, producing the destruc- 

 tive agent with definite velocities. In this respect, we may 

 compare the case to the case of destruction of the vestigial wing. 

 But here, as there, remains the problem of pattern. In the 

 vg-c&se, it was obviously a consequence of the structure of the 

 wing, prescribing the bed of spread of the lytic substance. In 

 the goldfish, the pattern seems to be imposed by the distribution 

 of lymph spaces from which the lytic action starts, a theory 

 actually assumed by Fukui. Goodrich points out that the 

 pattern in Oryzias and in Carassius is to be explained on a 

 different basis. There is no doubt that the latter is of the type 

 resulting from the interaction of a gene-controlled process 

 (production of the lytic substance) with what we termed the 

 conditions of the system. (We might apply the terminology 

 used before and speak also of the arrangement of the outlets for 

 the lytic stuff.) The principles obviously are always the same, 

 though the details may be very different. In the case of Oryzias, 

 we do not know the patterning agent, but no doubt it will fall 

 into one of the known categories. It may be added finally that 

 experiments in transplantation of scales (Goodrich and Nichols, 

 1933) show that the lytic action is limited to definite times in 

 development, as expected. 



In other fish species, more complicated patterns appear. In 

 Platypoecilus, for example, Gordon and Fraser (1931) have traced 



