282 FRAXZ SCHRADER. 



convert the derivatives of the polar nucleus to giant cells. It 

 is this point with which the present paper is concerned. 



The work was done on two species, Pseudococcus citri and P. 

 maritimus. 



CHROMOSOME COUNTS. 



The short statement regarding the origin of giant cells was 

 given as follows in my previous paper ('23) : "Later divisions of 

 the polar nucleus derivatives are subject to irregularities. Ap- 

 parently nuclear division is then very often or even generally not 

 accompained by cytoplasmic division, so that the two resulting 

 nuclei may lie side by side in a single protoplasmic area. At the 

 ensuing division, there may be an intermingling of the chromo- 

 somes evolved, or else a multiplicity of spindles. Possibly also, 

 cleavage cells nearing the edge may at times fuse with the deriva- 

 tives." 



This statement covers the problem only in a very general way 

 and is hardly definite enough to be regarded as a solution of the 

 complexities that are to be observed in the behavior of these 

 peculiar cells. The one point established is that the polar nucleus 

 derivatives are involved in some way in giving rise to the giant 

 cells. 



Without making any reference to this early period in the 

 embryology of Pseudococcus, Buchner ('21) in his work on sym- 

 biosis comments on the fact that the mycetocytes in the adult 

 contain a multiple number of chromosomes. This condition I 

 had also observed, but it was not until recently, when the embryol- 

 ogy was worked out, that the relationship between the giant 

 cells of the embryo and mycetocytes in the adult became clear 

 to me. Buchner apparently did not study the younger stages of 

 Pseudococcus and therefore did not observe the giant cells at all. 



The giant cells when first making their appearance in the egg 

 are marked by a peculiarity that was observed very early in my 

 investigations. This is, that although the number of chromsomes 

 they contain is clearly variable and always greater than the true 

 somatic number, it nevertheless varies only within very definite 

 limits. This led to a more careful examination of the chromo- 

 some numbers in these cells. 



A large number of chromosome plates was investigated. Un- 



