THE WILT DISEASE OF PIGEON-PEA. 31 
and the Indian species. The description and illustrations of the for- 
mer in Smith’s paper(2) are so full as to leave little to be desired 
for purposes of comparison. 
The perithecia are of the same tone of bright red in both cases 
as a comparison of plate Iof Smith’s paper, and plate I below will 
show. They are formed, in India as in America, on the roots below 
ground, rarely above ground, and are quite superficial. Smith 
gives their size as from 210 to 400, high, by 150 to 328, broad, 
mostly 250 to 350 by 200 to 300,. About the same limits have been 
found in the Indian forms, the most usual being from 250 to 350, 
high by 175 to 280, broad. The surface is almost smooth in both, 
beg slightly irregular from the projection of individual cells only, 
not groups of cells as m many Nectrias. The neck is usually short, 
that figured in plate II, fig. 9 (cotton Neocosmospora), where it is 
about gth of the total height of the perithecium, being perhaps the 
most common type. Sometimes it is longer (plate IJ, figs. 7 and 8), 
occasionally quite long as in plate I, fig. 3. Smith gives 30 to 40, 
as the usual length (or rather more than 1/10th of the total height), 
but says it may reach 80. I have found the greatest variability 
in this respect even in the same culture. 
The asci are numerous, cylindrical, shortly stipitate, the spore- 
bearing part measuring from 70 to 100 by 12,. They dissolve so 
as to liberate the spores within the perithecium when over-ripe. 
Paraphyses are present and consist of a chain of several rather large, 
irregularly oval, thin-walled cells, which are loosely connected. 
The figures in plate V, fig. 2, of Smith, and plate I, fig. 5, below, 
may be compared to show these characters. 
The ascospores are eight in each ascus, arranged in one row, 
globose to shortly elliptical, and with a brown thick wall of peculiar 
construction. This consists of an outer layer which is irregular in 
thickness and often wrinkled, giving the mature spore an irregular 
contour, and an inner layer of moderate and quite regular thickness. 
These two layers are not tightly fused together, as macerated or rup- 
tured spores show, Viewed in surface focus, the wall appears as if 
