Genic Control of Development 295 



first difiFerential division a change could have occurred in the genic 

 material which made it liable to endomitosis in subsequent cell genera- 

 tions. This does not sound very plausible, but, however that may be, 

 it cannot be asserted that determination via intranuclear change has 

 been proved in this case. 



We may derive the same conclusions from the work of Wiggles- 

 worth ( 1953 ) on bugs, though here possible polyteny has been neg- 

 lected. He showed that the diflFerentiation of epidermis cells into 

 sensory hairs is controlled by a substance in the epidermis needed for 

 determination of a bristle-forming center. Each center drains this 

 substance from the surrounding cells, as is shown by the fact that new 

 centers arise as far distant as possible from old ones. Some results of 

 burning experiments indicate that a certain concentration of the sub- 

 stance suffices for differentiation of gland cells but not of hair. Hair 

 differentiation (or dedifferentiation) of larval or adult type is under 

 the control of the juvenile hormone, all of which amounts to control 

 of cellular differentiation from outside the nucleus. 



In plants a comparable situation seems to exist. Geitler and his 

 school have made many contributions to this subject (see Geitler, 

 1954; Tschermak-Woess and Hasitschka, 1954). In general we may 

 say that the development of trichomes follows the model of the lepi- 

 dopteran scales. Polyploidy up to 256-ploid is found, while the cells 

 from which such large and specific structures originate remain diploid. 

 In view of the many examples found by Geitler and his school, it seems 

 that polyploidization is clearly connected with differentiation of spe- 

 cial large-celled organs. The meaning of this must be the same as in 

 the animal case, and the foregoing and following discussion applies to 

 both. 



In view of the importance of the subject, further discussion is 

 needed. In most of the well-known cases of endomitotic polyploidy 

 (including those in plants, the tapetum cells) we are dealing with 

 cells with a single intensive synthetic function. The gland cells, includ- 

 ing the dipteran saHvaries or the immensely polyploid silk-gland cells 

 of Lepidoptera, muscle cells, secretory intestinal cells, cells of Mal- 

 pighian tubules, are included in the list. It is certain that these cells, 

 polyploid by endomitosis, have stopped dividing but may grow to 

 very large size, which, in the cases studied (see Kurnick and Hersko- 

 witz, 1952), is proportional to the amount of polyteny. We may take 

 it for granted that polytene cells are incapable of regeneration of a 

 whole body, as, for example, leaf cells of Begonia can do. (The 

 nematodes with their cell-constant, highly polytenic cells are unable to 



