68 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
A second batch of seedlings will perhaps give 50 per cent., and a third 
trial 75 to 80 per cent. ; thus half results may be obtained with the third 
generation, and from 75 to 80 per cent. with the fourth. These succes- 
sive seedlings of the same variety are necessary to definitely insure the 
improvements obtained since the first generation. A careful selection 
must always be the principal aim of the raiser, because without 
that, far from succeeding in getting progress or improvement and 
fixing the definite success, the success itself degenerates and returns to 
the type. 
Discussion. 
Mr. F. W. Bursipas, M.A., V.M.H.: I should like to ask; Monsieur 
de la Devansaye whether he has ever obtained a hybrid between A. 
Andrianum and A. Scherzerianum. 
Monsieur DE LA DEVANSAYE: Never; nor have I ever heard'of one. 
