Theory of Plant Breeding 
cent. of a certain iris and which belonged to the inner 
row of three stamens which are usually absent. 
By breeding from these exceptional irises he, in 1894, 
found that 89 per cent. had extra stamens, and in 1895 
that 56 flowers out of 486 had six stamens, and also 
had all three petals changed so as to resemble the sepals.”” 
So that pure races may occasionally revert even to 
extremely distant ancestors. 
Another different question, which is also full of interest 
to all practical horticulturists, is the influence of the 
stock upon the scion, and wice versd. Mr. Daniel has 
published many observations upon this question and 
upon grafting generally. When, for instance, a perennial 
sunflower was grafted upon an annual one the stock 
became perennial, and other characters of the scion also 
made their appearance on the stock. 
Such characters as variegation, early flowering, fertility, 
resistance to cold and to parasitic fungi, have also been 
produced in the graft when attached to a suitable stock, 
It was by the use of American stock which were un- 
affected by the Phylloxera that the French and other 
vineyards were saved from destruction, but in this case 
the grafts were not (at least so far as the author is 
aware) influenced by the American stock. 
But these difficult questions must not be allowed to 
occupy too much space, and we cannot do better than 
close this chapter with an illuminating remark by the 
late Professor Weldon, “ There is probably no group of 
plants or of animals which cannot be directly modified 
by a suitable change in some group of environmental 
conditions,”’® 
1 Thomson. 2 De Vries. * Korschinsky, 
“ Garjeanne. * Molliard. ° Henslow. 
7 Arnott. * Blaringhem. * Von Wettstein, Henslow. 
10 Bateson. 11 Wilks. 12 and 18 Hurst. 
14 Saunders, 45 Davenport, 18 Chittenden. 
37 and ?8 Heinricher, Jones, 19 Weldon, 
306 
