THE MUTATED GENE 257 



of wasp queens, independent of the typical pattern of wasps, 

 is the product of a Liesegang ring formation with mechanical 

 evocation. 



There are rhythmical patterns which suggest at first sight a 

 Liesegang phenomenon. Henke (see 1935) calls them simul- 

 taneous rhythms and points out that they are characterized by 

 the absence of a center of diffusion and therefore must have a 

 different origin. Here belong the tiger and zebra stripes and their 

 like. Henke points to the fact that tiger and lion hybrids exhibit 

 a pattern that is a kind of compromise between the stripes of a 

 tiger and the rosettes of a young lion, which shows that the 

 different genes of these species control the same general process. 

 Many authors have tried to find an explanation for such patterns. 

 Haecker was the first to assume that rhythmic growth of the 

 skin was responsible, in so far as pigment would be deposited in 

 zones of intensive growth. This, of course, would refer the prob- 

 lem to the formation of a primary growth pattern. Goldschmidt 

 (1920c) for some time accepted this solution but was later led 

 to a different formulation when he had seen that a young tiger 

 embryo exhibited the whole later pattern in form of epithelial 

 papillae, which represented the Anlagen of the large hair (Leit- 

 haare). This same fact was found and elaborated by Toldt (1912) 

 in a series of papers. This pointed to a primary pattern forma- 

 tion in relation to the arrangement of hair follicles. Krieg (1922), 

 who returned to this subject, thinks of tensions within the skin 

 at the time of pattern determination ( = distribution of some- 

 thing) which makes the determining stuffs collect in definite 

 lines. This makes it possible, of course, by adding the rate 

 concept, to explain different patterns by different times of onset 

 of the action of determining stuffs. This view, indeed, accounts 

 well for the facts and has the advantage of permitting tests in 

 forms like dogs, cats, and cattle, where these patterns are con- 

 trolled by mutant genes, and other patterns by other mutant 

 genes. Such an analysis is still missing. Growth tensions, of 

 course, would then adequately explain a primary, general, and 

 simultaneous pattern, not showing any centers of origin. As such 

 tensions would be the automatic result of former developmental 

 processes, another type of automatic patterning could be added 

 to those already discussed. Henke (1935) points out that many 

 elementary processes of pattern formation might come into this 



