2 2 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
In some cases we have an unexplained reversal of symptoms in 
a given locality. Thus certain varieties of Aster, notably dumosus, 
were recorded badly mildewed at Wisley in 1911, but in 1914 and 
1 91 5 these were practically unaffected, whilst salicifolia, slightly affected 
in 191 1, was mildewed very much in 1914 and much in 1915. 
Incidence of Mildew (Erysiphe Cichoriacearum) on Asters (Michaelmas 
Daisies) at Wisley. 
Variety. 
Degree of Mildew. 
191X. 
1914. 
1915. 
Pink Daisy 
v. bad 
some 
little 
Porcelain 
slight 
much 
some 
Moonstone 
slight 
much 
little 
Snowdrift. 
little 
much 
some 
Daisy Peters 
v. bad 
little 
some 
White of the Dwarfs 
v. bad 
little 
some 
salicifolia 
slight 
v. much 
much 
Janus .... 
slight 
much 
slight 
Thirza .... 
slight 
much 
slight 
dumosus . . . . 
bad 
none 
none 
Grace .... 
unrecorded 
none 
much 
Ariadne .... 
v. bad 
v. much 
some 
Apollo .... 
bad 
v. much 
much 
Berenice .... 
v. bad 
much 
some 
canus .... 
v. bad 
v. much 
much 
A similar happening was observed amongst Pears affected with 
scab caused by Venturia pirina from observations made at Wisley in 
1914 and 1915. The significance of phenomena akin to these occurring 
among plants propagated vegetatively, such as the sugar-cane, has 
been ably discussed by Barber and by Butler in India, and by 
Harrison, Stockdale, and Ward in the West Indies. 
There seems no prospect that the use of preventive and curative 
remedies can be to any extent abandoned, and especially since 
susceptible varieties of fruit, vegetables, &c, are commercially prized 
through possessing some favourite quality of colour, flavour, fragrance, 
and form. We can, however, abandon their improper use and direct 
our efforts to improve their utility. Some discredit has been thrown 
on the practice of spraying and the use of bouillies, owing to the 
conflicting results obtained ; such results will inevitably accrue from 
an insufficient knowledge of the nature and origin of the diseases or 
epidemics treated, and furnish another cogent argument for increased 
investigational activity. I will give a case from past experience which 
exemplifies the kind of misconception responsible for error in many 
more widely planned experiments. Several rows of potatos in 
a large potato field in England were sprayed with the object of 
impressing the farmer with the beneficial effect of spraying against 
potato blight (Phytophthora infestans). At the time of lifting, how- 
ever, it was found that although the foliage was healthy the tubers 
formed by the sprayed plants were badly diseased, and to the same 
extent as those of the unsprayed plants (foliage also healthy) . The 
