10 SCIENCE PROGRESS. 



The above cases show that the former influence, mz., 

 that of the older-estabhshed species, is the more effective of 

 the two. The sexual prepotency of the male 6'. pavonia 

 counts for less than the specific prepotency of the female 5. 

 spini. The highest effect of course is produced when the 

 two influences concur, as in the hybrid of S. spini S and 

 6'. pavonia ? . 



A further result of the experiments is that while no 

 female hybrid was proved to be fertile, there are undoubted 

 cases of fertility in male hybrids. This has been shown by 

 crossing with the females of both parent species, and in one 

 case with a female of a third species [S. boinieinanni $ and 

 6". pyri ? ). 



Hence hybridisation is not necessarily a merely tran- 

 sitory phenomenon. There seems to be no reason why a 

 male hybrid should not propagate by " back-crossing " under 

 natural conditions ; and since the product of this kind of 

 crossing is not found to show a complete reversion to the 

 type of the female parent, it is possible that the existence of 

 various intermediate forms in such genera as Melitcsa, 

 ZygcEiia and Agi^otis may be accounted for in this manner. 

 Cases of simple pairing between distinct species of the two 

 former genera have been observed by the author in nature. 



The above conclusions with the facts on which they 

 rest, though much condensed from the original, have been 

 stated as far as possible in Dr. Standfuss's own words. A 

 few brief comments seem called for. 



(i) The rule laid down by Standfuss as to the pre- 

 potency of the phylogenetically older species is probably 

 another expression of the fact so clearly established by Dar- 

 win ^ that hybridisation frequently leads to reversion. It is 

 significant that Standfuss considers the hybrid form S. 

 emilicB as partly reproducing an ancestral stage in the 

 history of S. pavonia rather than the form of that species 

 at present existing. 



(2) It is well known that when plant hybrids are crossed 

 with one of the parent species, considerable variability may 



'^Animals and Plants ttnder Domesticatio7i, 1868, vol. ii., p. 254, etc. 



