40 E. J. BUTLER. 
Here, again, the wilt was typical and virulent. All the plants 
(nearly 50) in the moculated pots were dead by the early part of Sep- 
tember, the control plants remaiming perfectly healthy. 
Series XVI. 
The cultures were obtamed as in Series XV, and were grown on 
carrot, onion, plantain and rice. 
The plants were grown ina small plot nm my garden, 8 by 4 feet 
in size, the seeds being steeped in formaldehyde as before. 
The moculations were made as in the last two series. 
No. of | Date of sowingand | ; 
plants. Treatment. aaulation: Result. Remarks. 
Gontents | Inoculated with|Sown and inoculated | Wilt appeared. Ist death on 
of a plot. Cephalosporium | 1-7-08. | 10-8-08. A large 
8x4feet. from pigeon-pea. | |/number were 
dead by 26-10-08. 
Ditto Not inoculated | Sown on 1-7-08. No deaths up to 
(control). March, ’09. 
The two plots were side by side, separated by a small trench 
about a foot in depth. It is somewhat remarkable that there was 
no spread from the moculated to the control part by aerial conidia, 
as other experiments seem to show that such a spread occurs not 
uncommonly in nature. 
Series XVII. 
The cultures used were the same as in the last series. 
The plants were two clumps over a year old, growmg in open 
soil in my garden, and were the contents of pot No. 3 (inoculated 
previously with Neocosmospora) and pot No. 7 (not previously 
inoculated) of Series V above (page 10). These had been cut back 
when about 9 months old and had given a second vigorous growth, 
bemg still perfectly healthy. 
The inoculations were made by scraping away the soil down 
to the finer roots and burying part of the culture in contact with the 
latter 
