THE BEHAVIOUR OF INDIVIDUAL CHROMOSOMES 85 



"hypoploid for") factors contained in the chromosome which has been 

 lost. If the lost chromosome contained dominants masking the presence 

 of the corresponding recessives in the sister chromosome, these reces- 

 sives will now be able to show, perhaps with exaggeration. 



Mosaics in which an X chromosome has been eliminated from a 

 female show male characters in the parts from which the second X is 

 missing. They are known as gynandromorphs.^ The eliminations seem 

 to be most usually in one of the earliest segmentation-divisions of the 

 egg, often at the first, and gives rise to a gynandromorph with approxi- 

 mately one half of the body male and one half female.^ Animals mosaic 

 for sex or other characters are known in birds and mammals^ as well as 

 insects. By bringing tissues of genetically different constitutions into 

 intimate contact with one another they provide opportunities for 

 investigating the mutual interaction of the tissues during development, 

 a subject which will be discussed more fully later (p. 177). Here it is 

 sufiicient to remark that in insects the remarkable thing is not the 

 interaction of the different tissues but their lack of interaction, each cell 

 developing the characters corresponding to its own nucleus with little 

 reference to the genetic constitution of its surroundings. For instance, 

 gynandromorphs in the parasitic wasp Habrobracon, with a head of one 

 sex and gonads of another, are said to behave purely according to the 

 type of their cephalic ganglion;^ but this praiseworthy effort to make 

 the sex Ufe subservient to the dictates of reason seems to be even less 

 likely than usual to succeed in this case. 



If the mosaic is of such a kind that each cell lacking the eliminated 

 chromosome can be recognized, it may enable one to distinguish a 

 certain group of cells as necessarily derived from a single cell in which 

 the elimination took place. This would make it possible to trace the 

 cell lineages of insect development, which would be a considerable 

 refinement of the description of insect development as it is known at 

 present. So far only the explanation of gynandromorphs in Droso- 

 phila has been fully worked out on these lines^ but factors have been 

 found which give a high percentage of chromosome elimination at late 

 stages of development,^ and the resulting small mosaic patches may 

 enable cell Hneages in insects to be described in greater detail.'^ In some 

 plants ring-shaped chromosomes are known which suffer frequent 



^ Cf. Morgan and Bridges 19 19. '^ Cf. Sturtevant 19296. 



^ E.g. mouse, see Dunn 1934. " Whiting 1932a. ^ Parks 1936. 



® It is probable that these should be re-interpreted as due to somatic crossing- 

 over (p. 373). ' Bridges 19256, Noujdin 1936, Sturtevant 19296. 



