PA'I'EKNAL CHARACTERS IN ECHINOID HYBRIDS. 
347 
up to the present no hybi-ids liave been rnised to sexual 
maturity, any suggested Mendelian explanation must remain 
unconfirmed. Nevertheless we would suggest that the follow- 
ing liypothesis might be put forwaid to express the purely 
maternal inheritance. Suppose that the presence of a pair of 
posterior ciliated epaulettes (P) and the absence of the same 
(p) are a pair of allelomorphic characters. Then the germ- 
cells of E. acntus vvill all have P and those of E. miliaris 
will all have p. Suppose, moreover, that in hybrids when P 
comes in through the male it is recessive to p, but when 
through the female it is dominant, 
then E . acntus ? X f recessive p through tliej* 
E. miliaris d* receives t dominant P through the ? 
and E. miliaris ? X f recessive P through the d 
E. acntus d I'eceives ( dominant p through the $ j epauleU^s^'^^”'^^ 
If this hypothesis were true and segregation took place, the 
Ej generations from both i-eciprocal crosses would have to be 
composed of larvte in equal numbers with and Avithout the 
posterior epaulettes. If, on the other hand, there were no 
segregation, all the F3 individuals would be alike and similar 
to the Pi hybrid from which they were bred. 
We have explained above that two of our crosses have this 
year shown a complete reversal in the inheritance of the 
])ai*ental characters. While in hybrids between E. escu- 
lent us or E . ac u tus and E . miliaris, in which the latter 
was used as the Jiiale parent, the characters were maternal as 
formerly, when E. miliaris was used as the female the 
hybrids this season usually resembled the father. Formerly 
they had been invariably mateinal. This complete change in 
heredity from one year to another seems to be without parallel 
in hybrids other than those amongf Fchinoderms. Using the 
same notation as above we should have to suppose that this 
year P behaves as a simple dominant over p. 
As we have emphasised above, previous investigators in 
Echinoderm hybridisation have used charucters wliich were 
not definite enough to give unquestionable results. Never- 
the hybrid hav- 
ing posterior 
epaulettes ; 
the hybrid, hav- 
