Mutations and Evolution. 
137 
to be homosynaptic. After heterosynapsis the reductions will be 
XX—Y or X—XY. From the former the eggs after maturation 
will be |XX and £Y; from the latter, JX and XY. Also from 
homosynapsis, the reductions will be X—XY or XY—X, since the 
unmated Y may remain in the egg or enter a polar body, and the 
eggs will therefore be -JX and JXY. Hence 4 classes of eggs result: 
XX and Y 4 % each, of single origin; X and XY 46% each, of 
composite origin. Therefore if such an XXY female having 
vermilion eyes is crossed by a wild male (red eyes), there will be 8 
classes of zygotes, as shown in the diagram (Fig. 2). 
Of these eight classes, (2) are not exceptions but would 
produce exceptions in the next generation, (4) are produced by the 
reverse of the ordinary method, i.e., by the union of a Y egg and an 
X sperm. While themselves exceptions, they are ordinary males 
and can neither produce exceptions nor transmit the power of 
doing so, (6) are not exceptions, but some of their daughters will 
get the extra Y (from XY sperm -f X egg) and so produce 
secondary exceptions. It was found that homosynapsis occurs 
much oftener than heterosynapsis, but otherwise the chromosome 
distribution is according to chance. That matroclinous exceptions 
had developed from fertilised eggs was shown by the presence of 
paternal characters derived from other chromosomes. 
In XYY males, synapsis may be of the XY or YY types. If 
these are simply by chance there should obviously be twice as 
many XY as YY synapses. The synapsed chromosomes disjoin 
and the other one apparently goes equally to either pole. Hence in 
XYY males 4 classes of sperm will be produced, X and YY from 
heterosynapsis (small classes) and XY and Y from homosynapsis 
(twice as large and from two sources). These four classes of sperm 
fertilizing an ordinary female would give the following results : 
Sperm X YY XY Y 
Eggs X 
XX XYY XXY XY 
? $ ? $ 
Half the daughters will therefore be XX and half XXY, while 
half the sons will be XY and half XYY, again producing exceptions. 
Primary non-disjunction in the female, or failure of the sex 
chromosomes to separate in the egg, occurred 12 times in breeding 
experiments involving 20484 flies, or 1 in 1700. The fact that XO 
males are sterile shows that the Y chromosome has some function 
in spermatogenesis, if only to serve as a balance wheel. 
