141 
both the saltant and the original to the same plate so that they may have 
the same environmental conditions, the identical quantity and quality of 
agar, and by growing close together may render comparison easy. Notes 
on the origin and subsequent behavior of the saltants were made under the 
serial number, ard the transfers were designated by additional numbers 
Thus, M98-7 refers to saltant No. 98, transfer 7. 
CHARACTERS OF SALTANTS AS SHOWN IN TRANSFERS 
General appearance.—The colonies of the saltant and of the original 
when grown on the same plate were usually so strikingly different in gen- 
eral appearance (Fl. XXII, XXIII) that a mere glance sufficed to give 
the impression that they were colonies of two different species. This dif- 
ference in general appearance is, on analysis, referable to one or more of 
the individual differences mentioned below. 
Fate cf linear growth.—Frequently the saltant was of much slower 
growth than the original, resulting in an M colony of much less diameter 
than that of the O colony, being often less than half of it (see two ex- 
amples: one given in Pl. XXII and one in Pl. XXIII). In some instances, 
however, the M colony grew faster than the O colony. 
Cemidial production.—Frequently the M colony, especially when slow- 
growing, was much more productive of conidia than the O colony, so much 
so as to give the colony a decidedly perceptible darker color. In several 
instances, however, the M colony was of the opposite character, producing 
few conidia or, in some cases, going to the extreme of appearing to pro- 
duce none at all. Generally speaking, rate of linear growth was in inverse 
ratio to that of conidia-production; while those saltants that were pale and 
possessed much aerial mycelium were usually of rapid linear growth and 
low conidial production. 
Conidial clusters-—Some saltants varied strikingly from each other 
and from the originals in the mean number of conidia borne per conidio- 
phore. 
Conidial length, breadth, septation, and shape.—TYhese characters, as 
evidenced by casual observation or by a study of graphs and the data 
derived from them, are shown to be strikingly different in various saltants. 
For clearness I present in this connection records concerning only a few 
saltants, giving graphs and data for others later. 
Graphs of conidial length of saltants M35, M36, and M40, those which 
show greatest deviation from originals in this regard, are given in Fig. P 
with the essential data. It is to be observed that the modes of M35 and 
