776 
Journal of Agricultural Research voi. xxvii, No. i 
CONIDIOPHORES 
In the case of the conidiophores, for example, a study of optimum 
nocturnal material shows clearly that the conidiophores are larger, more 
extensively branched, and structurally more complex and elaborate, than 
most of the illustrations and descriptions indicate. The first description 
of the conidial stage of this species was published by Schroeter (22) and * 
the first illustrations were shown by Fischer (5, fig. 71). Together these 
present a vivid characterization of the conidiophores as decidedly short 
(about iooju) and thick (about 12 fi) with few stubby branches bearing a 
relatively small number of conidia (about 15). No appreciable departure 
from this characterization is found in the publications by such subse¬ 
quent investigators in Europe as Berlese (r), Malbranche and Letendre 
(14) , Massee (15, pi. 1, fig. 16) , Saccardo (jp), et al.; or such in America as 
Saunders {21, pi. 16 , fig. 4), Farlow (4), Wilson (jj), etal.; or in India 
as Butler (2, fig . < 5 , 7), and Kulkarni ( 12 , fig. 1-3). In Japan, how¬ 
ever, Shirai (24, fig. 16, 17) described the conidiophores as much larger 
(100 to 240/z by 12 to 19/x) and with a somewhat more elaborate branching 
system; but the fact that he found conidia of two strangely different 
size-classes (24 to 28.8/x by 16.8 to 19.2/u, and 38.4 to 57.6 m by 19.2 to 
24M) arouses the suspicion that he was not dealing with 5 . graminicola 
alone, and to some extent invalidates his characterization of the species. 
The writer, after studying the conidiophores of Sclerospora graminicola 
in the progressive stages of their development under optimum conditions 
at night, is convinced that the descriptions and drawings of the investi¬ 
gators just mentioned are based on nontypical, poorly developed speci¬ 
mens, the last belated stragglers of the nocturnal production, caught by 
the morning sun before they had developed conidia, and dried to a con¬ 
dition still less typical and favorable for study. In contrast to such 
dried specimens, material scraped from Setaria leaves at the time of 
optimum nocturnal production shows conidiophores that are much 
larger and better developed than we had been led to believe characteristic 
of 5 . graminicola , and which approximate, in this respect, those of such 
luxuriant species as the conidial Sclerosporas of the Orient. If, for 
example, one compares the accompanying figures of 5 . graminicola (PI. 
2, I, K, O) with those of 5 . spontanea (jo, pi. 7g } fig. A ), and S. philips 
pinensis (29, pi. 24 , fig. O , pi. 23, fig. B) all of which are of very nearly 
the same magnification, it is obvious that they show a general resemblance 
which is not even suggested by previous illustrations. 
After comparing them carefully, however, the writer finds that even 
the largest and most elaborate conidiophores of S. graminicola differ 
markedly in certain essential features from the conidiophores of such 
oriental species as S. philippinensis and 5 . spontanea. 
First, the total length in these Philippine species is as a rule much 
greater (260 to 400^) than in the case of 5 . graminicola , even the most 
luxuriant nocturnal material of which has a length of about 150^ with 
occasional extremes as low as ioom or as high as 200 fx. 
Second, the branching system also shows differences which are much 
more qualitative and absolute than are mere distinctions in size. In the 
Philippine species, for example, usually three,* sometimes two or four, 
primary branches of approximately equal size and extent of development 
all spread out at angles of about 45 0 or less from the main axis in very 
close succession; and all are of equal rank—no one of them being either in 
direction, position, or extent of growth more to be considered a continua- 
