Structural Differentiation of the Nucleus 117 



owe their origin to endomitoses. Such somatic polyploidy is common 

 (see Oksala, 1939). Smith (1941) found the chromosome number 

 to be doubled in wing buds of sawflies (Hymenoptera) , i. e., to be 

 2n in males and 4n in females. In the ovariole wall it was octoploid 

 and in neural cells 20-30n. In leguminous plants, Wipf and Cooper 

 (1938) found that dividing cells in the root nodules, which are 

 involved in the process of nitrogen fixation, are characteristically 

 tetraploid. 



The clearest case of endomitosis in plants occurs in spinach, 

 2n = 12, where somatic polyploidy has long been known, but its 

 origin due to endomitosis was shown only recently by Gentcheff 

 and Gustafsson (1939). They find that in the periblem cells of 

 roots from germinating seeds, the nuclei have as many as 96 

 chromosomes. This polyploidy is correlated with the presence of a 

 large amount of storage products in the periblem cells. The chromo- 

 somes are split at the beginning of prophase. At metaphase they lie 

 in pairs and do not separate. After a resting stage they are present 

 in the double number and unpaired. Levan (1939) obtained similar 

 results in auxin-treated cortical cells of spinach. In Allium, auxin 

 produced endomitosis differing only in that the kinetochores were 

 delayed in their division relative to the arms, and "diplochromo- 

 somes" were therefore present at metaphase. In these experiments 

 with auxin, cellular enlargement precedes the increase in nuclear 

 volume, and endomitosis appears to be initiated thereby. In most 

 of the other cases cited, chromosome or chromonema multiplication 

 precedes and apparently initiates the increase in nuclear volume. 

 This is also the case in polyploidy produced experimentally with col- 

 chicine. Colchicine apparently inhibits the centrosome or spindle 

 activity and sometimes the reproduction of the kinetochore, but not 

 that of the chromonema. Levan obtained cells with as many as 1,000 

 diplochromosomes, i. e., chromosomes divided except at the kineto- 

 chore, in Alliinn which normally has 16. 



The giant chromosomes of the salivary gland nuclei of Diptera 

 are now generally, though not universally, conceded to be bundles 

 of chromonemata. They appear to result from successive endomitoses 

 which apparently differ from the foregoing cases chiefly in that the 

 chromonemata remain associated, instead of falling apart and 

 constituting separate chromosomes, and that homologues become 

 tightly paired. These differences may both well be concomitants 

 of the closer association which exists between homologous chromo- 

 somes in all cells of the Diptera— Metz (1916) showed that in poly- 



