Chromosomes and Genes 189 



formation on this overlap effect to be certain that it is not different 

 from the effect within the segments. However this may be, the 

 problem remains why pattern changes within the segments are always 

 allelic to each other and those in different segments are not; though 

 the facts of overlap show that the latter is not completely true, but 

 that there is a tapering off of the allelic features beyond the end of 

 the segment. Some authors are content with the answer that the 

 segments are the genes and therefore allelic. Those who are not 

 satisfied with this answer may try to find a solution in terms of genie 

 action. The chemical constitution of each section may be such that its 

 direct products control a master set of reactions leading to an end 

 product which is necessary for a certain developmental step. Pattern 

 changes within the segment do not affect the master reaction as such 

 (just as many stereoisomeres of an active substance might not change 

 its activity). Some other repatterning might impair the master re- 

 action and therefore be more or less lethal (just as some stereoisomeres 

 of an active substance prevent the activity; good examples are found 

 in Ruzicka's work on testosteron isomeres ) . But most repatterning will 

 produce secondary effects in speed of reaction, quantity of the 

 products, and the like, which become visible in the slight departures 

 in the end product, the visible mutant effect. As allelism relates to 

 this end product, it is the expression of the one untouched master 

 reaction. 



Thus far this interpretation would also fit a new definition of the 

 gene as a chromosomal segment within which any change at any point 

 produces an allelic mutant effect and within which, also, crossover 

 breaks may occur. Such a definition would make the gene a more or 

 less large, visible element with a visible substructure. The classic 

 beadlike gene would remain and not only change in size, complication, 

 and divisibility, but also in lack of definite limits (the overlap effect). 

 However, the facts have forced us to turn from an independent, 

 homogeneous corpuscle to an indefinitely limited part of a whole 

 having a typical serial pattern which alone allows for normal function. 

 The logical consequence is that the section, acting as a small field, is 

 itself only a part of a larger field or different fields, also overlapping 

 ones. In terms of the model of function just presented, this would 

 mean that the "master reaction" controlled by the small field, the 

 segment, is related to the still more primary effects of the larger field, 

 as the pattern change within the segment is related to the whole 

 segmental action. This means that a pattern change — mutation — 

 within any segment would not much disturb the action of the next 



