504 
[,. DONCASTER AND J. GRAY. 
Summary and Discussion. 
Our work was undertaken primarily in order to discover 
whetlier in the crosses described there was any systematic 
elimination of chromosomes, such as have been found by 
Baltzer in crosses between Sphaerechinus and Strongylo- 
centrotus, which might be correlated with the facts 
described bj^ Shearer, De Morgan and Fuchs in their work 
on the hybrid plutei. Baltzer found that in the cross 
Sphierechinus ? x Strongylocentrotus cJ no elimi- 
nation of chromosomes occurs, and tlie hybrid plutei are 
intermediate between the parental types. In the converse 
cross (which is successful iu a very small percentage of 
cases), sixteen chromosomes, regarded as belonging to the 
male (Spliaerech inns) parent are eliminated, and the hybrid 
plutei, when exceptionally they can be reared, are of the 
maternal type. In our cases. Shearer, De Morgan and Fuchs 
have found, in the years preceding 1912, that in crosses 
between esculentus or acutus and miliaris the plutei, in 
certain characters at least, were of the maternal type, which- 
ever way the cross was made. It seemed possible, therefore, 
that we might find elimination of chromosomes in the manner 
described by Baltzer. 
For several reasons the work described above has followed 
somewhat different lines from those expected at the beginning. 
In 1911, the only cross with miliaris which was obtained 
before the season was too far advanced was acutus ? x 
miliaris In the hybrid eggs we found that some of 
the chromosomes developed vesicles, but no other elimination 
occurred except the possible non-division of certain chi-omo- 
somes, about which we are uncertain. The vesicle formation, 
which appears to a small extent in eggs of acutus ? x 
miliaris (^, is much more clearly shown in the cross acutus 
? X esculentus c? , species in which the characters of the 
plutei are so similar that the hybrid cannot be distinguished 
with certainty from the pure forms. We have therefore been 
