Mar. 8, 1924 
Nocturnal Production of Sclerospora 
777 
tion of the main axis than any other. As a result of this also, the conidia 
arising from the branch tips are arranged approximately in a hollow 
hemisphere. In 5 . graminicola , on the contrary, one of the primary 
branches stands out more or less obviously as a continuation of the main 
axis (PI. 2, I, K, L, O, Q) both in direction and in extent of growth. 
From this continuation of the main axis other main, and secondary 
branches grow out at irregular intervals, usually at angles of 45 ° to 90°. 
As a result, the conidia at the ends of the branches lie more frequently 
in irregularly disposed groups (PI. 2, I, K, L, P) than in an approxi¬ 
mate hemisphere (PI. 2, O, Q). 
Third, the sterigmata of 5 . graminicola , as a rule, are shorter and more 
broadly bottle or tenpin shaped (PI. 2, N, T, V, W) than are those 
of the Philippine Sclerosporas. The shape, however, is somewhat de¬ 
pendent on position, single sterigmata at the tips of the branches (PI. 2, 
U) being more elongate than those borne in groups of two, three, 
or four (PI. 2, T, J, W). Moreover, the sterigmata, which in the 
Philippine Sclerosporas almost invariably continue the direction of the 
branch tip that bears them, may stand out from the branch tip even at 
right angles in S. graminicola (PI. 2, I, J, L). Finally, the lower 
part of the conidiophore of S. graminicola is not usually marked off by 
a cross wall into a basal cell or foot cell that is distinct from the superior 
portion of the main axis. Occasionally such basal cells are encountered 
(PI. 2, K, X, Y); but this condition is the exception; and hence is 
in distinct contrast to the Philippine species in which it is a character¬ 
istic feature. Occasionally, also, the base is distinguished by an incom¬ 
plete transverse septum (PI. 2, K) or by a decided thickening of its 
longitudinal wall (PI. 2, A, P, Z). 
All these distinctions—size of the conidiophores, character and extent 
of the branch system, form and direction of the sterigmata, and extent 
of septation or thickening at the base of the main axis—are matters of 
degree which should be expressed quantitatively to facilitate compari¬ 
sons. Yet, even when considered as qualitative differences, they show 
clearly that the conidiophores of S. graminicola, though they may approxi¬ 
mate those of such typically conidial oriental forms as S', philippinensis 
and 5. spontanea in luxuriance and general appearance, are indeed dis¬ 
tinct from them. 
conidia 
It is the conidia themselves, however, that are the most distinguishing 
feature of Sclerospora graminicola . These bodies differ markedly from 
those of other Sclerosporas in size, shape, structure, and germination. 
The size of the conidia varies, and for adequate presentation requires 
quantitative expression based on measurements of large numbers of 
conidia. Moreover, to be ideally satisfactory the conidia should be caught 
on the slide in dew when snapped off from the conidiophores at maturity 
and measured at once. Unfortunately, the writer was not able in this 
case to make all the measurements under these ideal conditions as he did 
for the Philippine Sclerosporas. Rather, most of the measurements were 
made from material scraped from abundantly productive leaves at 3 a. m., 
then killed by Flemming’s weaker solution, and mounted in dilute glyc¬ 
erin and eosin. A comparison of measurements made thus with the rela¬ 
tively few which the writer had opportunity to make from fresh material 
indicates that if all the measurements had been made under ideal con¬ 
ditions the modes of length and of diameter probably would have been 
83798—24 - 5 
