E. HORTON AND E. S. SALMON 273 



tion exerted a slight checking action on the mildew-patches on two of 

 the five sprayed leaves ; on the remaining three leaves the mildew-patches, 

 on the fourth day after spraying, were as ''powdery" as those on the 

 "control" leaves (unsprayed). The solution at this strength, then, was 

 practically non-fungicidal. 



Experiment 19. Tricalciiun arsenate (0-076 per cent. AsjOj). By the 

 fourth day after spraying the mildew-patches on all the sprayed leaves 

 (five) were barren and apparently dead; on the "control" leaves (five), 

 sprayed with 1 per cent, calcium caseinate, all the mildew-patches were 

 very vigorous and densely " powdery." By the sixteenth day it was clear 

 that the solution had been completely fungicidal. No trace of "scorch- 

 ing" appeared anywhere on the leaves. 



Experiment 31. Tncalciinn arsenate (0-02 per cent. As^Oj). By the 

 third day several of the mildew-patches on four of the (seven) sprayed 

 leaves had developed fresh conidiophores at the edges of the patches — 

 a clear indication that the solution was not quite fungicidal ; on three of 

 the leaves the patches were all barren. At this date all the mildew- 

 patches on the seven "control" leaves, which had been sprayed with 

 1 per cent, calcium caseinate, were very vigorous and densely " powdery." 

 At the end of the experiment, i.e., on the twelfth day, the condition of 

 the mildew on the sprayed leaves was as follows: leaf (1), (2), (3), at the 

 4th, 5th, 6th nodes; clustered conidiophores round the edges of all the 

 patches, while the nujcelium was barren and probably dead at the centre 

 of each patch. This condition indicates that the solution used was of a 

 not quite fungicidal strength. Leaf (4) at the 3rd node, and leaf (5) at 

 the 5th node; the mildew-patches now bearing densely cluistered "pow- 

 dery" conidiophores at their edges — indicating that the solution was 

 almost non-fungicidal. Leaf (6) and (7), at the 6th and 7th nodes: many 

 of the patches remained permanently barren and were probably killed; 

 a few developed weak scattered conidiophores at their edges. Here the 

 solution proved almost but not quite fungicidal. In previous experi- 

 ments (2) it had been found that a "powdery" patch of mildew is easier 

 to kill when growing on an older leaf than on a younger leaf. 



Experiment 34. Tricalcium arsenate (0-01 per cent. AsgOg). By the 

 seventh day only a slight checking action on the growth of the mildew 

 was evident, most of the patches now bearing densely clustered, more or 

 less "powdery" conidiophores. It was clear that the solution at this 

 strength is practically non-fungicidal. 



19—2 



