294 Action of the Genetic Material 



much-quoted example is the differentiation of egg and nurse cells in 

 Drosophila (Painter and Reindorp, 1939), in which nurse cells, which 

 will not divide any further, develop polytene nuclei by endomitosis. 

 The best material for a discussion is, however, found in Henke and 

 students' (Henke and Pohley, 1952) subtle and painstaking work on 

 the differentiation of the scales on the wing of the flour moth. Figure 

 17 presents Henke's diagram of the main points. Two' types of dif- 

 ferentiation must be distinguished. One is the final development of the 

 scale itself from a stem cell, which has a definite lineage going back 

 to a primordial wing cell. This differentiation takes place (top of the 

 diagram) when a scale-forming stem cell first divides into two, one 

 of which degenerates. The surviving stem cell divides again into two, 

 one of which is the final scale-forming cell which grows enormously 

 to form the scale. The other is the sheath-forming cell, which remains 

 small and develops into the sheath at the base of the scale. The latter 

 remains diploid in smaller scales, but becomes tetraploid in larger 

 ones. However, the actual scale cell becomes polyploid by endomitosis 

 up to 32 n. This same is true in principle for all types of scales, of 

 which four are represented in the diagram. But — and this is the 

 second feature — the amount of polyploidy is different according to 

 the size of the scale types ( also, large and small sheath cells are tetra- 

 ploid and diploid, respectively). In the diagram 8- to 32-ploid nuclei 

 are represented for the different scale types. 



Our problem now is whether differentiation is the result of intra- 

 nuclear differentiation of the genie material, in the present case by 

 endomitosis. This would mean that the result of endomitosis is a con- 

 dition of the genie material, different from normal, which causes the 

 differentiation. If we take as our example only the last steps (on the 

 right in the diagram), the differential division occurs so that one 

 daughter cell develops a polyploid nucleus, and this cell grows, giving 

 rise to a scale; the other cell stays diploid ( or becomes tetraploid ) and 

 remains small to form the sheath. At its face value, this seems to be a 

 beautfful case of intranuclear differentiation as a cause of divergent 

 determination, but this conclusion cannot stand closer scrutiny. Since 

 the last division always gives one sheath and one scale cell, the differ- 

 ential determination has already been accomplished before growth 

 and endomitosis begins. If we go back in the cellular pedigree, we 

 find the same situation at each differential division. At the time of the 

 first separation (left in the diagram) of epithelial and scale stem cells, 

 the future scale type and therefore the amount of endomitosis already 

 has been determined. Someone might retort that at the time of the 



