^92 R . Ruggles Gate$. 
nuclei (presumably the female) in the fertilized egg; (ii.) the apparent 
presence of only 7, the x number of chromosomes, in the cell- 
divisions of the embryo; (iii.) the smaller size of the embryo and 
its cells and nuclei in biennis x muricata than in biennis, which 
would be anticipated according to Boveri’s law. Nevertheless, there 
are various difficulties, and the results of a fuller study of the 
subject will he awaited with interest. A number of other questions 
remain to be solved. If the hybrid embryo begins its divisions with 
only 7 chromosomes it is not probable that this number persists 
throughout the sporophyte, and Goldschmidt believes he has found 
evidence that the 2x number is restored in the later mitoses of 
older embryos. This introduces difficulties, however, for it is 
scarcely probable that all the cells would double their chromosome- 
number simultaneously, and if both types of cells continued we 
should anticipate as a result various distortions or changes in the 
shape of organs, for we should have different chromosome-numbers 
in different parts of the individual. The manner of chromosome 
pairing in meiosis would also be of interest, though we might 
anticipate the formation of seven pairs by analogy with the 
chromosome behaviour in the tetraploid mutant gigas. If more 
critical study sustains Goldschmidt’s foreshadowing, we shall be able 
to add one more to the number of cases in CEnothera where cytological 
research has furnished the key to certain otherwise obscure or 
inexplicable hereditary phenomena. 
It is worthy of mention that these reversions to one of the 
grandparents in double reciprocal crosses were predicted by Giglio- 
Tos (22) on generalized highly theoretical grounds concerning the 
structure and arrangement of protoplasmic elements, but it is highly 
improbable that this type of hereditary behaviour will be found to be 
common even in the genus CEnothera itself. 
If Goldschmidt has correctly ascribed these results tomerogony, 1 
then a similar explanation may be applied (as that author has 
already done) to the various cases of “ false hybrids,” such as the 
well known results of Millardet with strawberries. In this case the 
hybrids were found to be all purely paternal in character and to 
breed true. The converse condition known in various Orchids, in 
which the hybrids are purely maternal in type, may conceivably 
result from the degeneration of the male nucleus after entering the 
egg or, as has often been suggested, from stimulation of the egg to 
parthenogenetic development by the presence of the pollen-tubes. 
Another series of interspecific crosses has been made by Davis 
(7, 8, 9). He chose races of O. biennis from Massachusetts and of 
0. grandiflora from Alabama for the parents of his crosses, in the 
hope of producing 0. Lamarckiana. But it must be said that from 
this point of view the results have not been successful. Hybrid 
forms were obtained, some of which showed a certain amount of 
crinkling of the leaves, but that is only one of many features of 0. 
1 Since the above was written, Renner (“ fiber die angebliche Merogonie 
der CEnotherabastarde,” Ber. d. deutsch. bot. Ges,, Bd. 31, pp. 334-335, 1913) 
has re-investigated the subject in 0. muricata x 0 . biennis, 0 . biennis x O. 
Lamarckiana and 0. Lamarckiana x O. biennis. He finds the usual double ferti¬ 
lization in all cases, with 2.x chromosomes in embryo and endosperm. Re¬ 
examination of Goldschmidt’s preparations leads to the same result, so it must 
be concluded that there is no evidence for the theory of merogony in these 
hybrids as expressed by Goldschmidt. 
