145 



conidial production and rice versa. Correlations observed are indicated 



as follows: 



Slow linear growth* >high conidia-production 



Much aerial mycelium >low conidia-production 



Pale colony* ^rapid growth 



Thickening of conidia ^shortening of conidia 



Pale colony >low conidia-production 



Clumping of mycelium low conidia-production 



1 he differences in colony-color and growth-rapidity here noted, are 

 much like those described by Edgerton (51) in Glomerella plus and minus 

 strains. Crabill (36) notes also a correlation in that his minus strains were 

 always of more rapid growth than the plus strains. 



TENDENCIES IN SALTATION 



Saltants showing very low conidia-production, verging on sterility, 

 coupled with paleness of colony, occurred with the greatest frequency. A 

 type with increased conidia-production and of slow growth was next in 

 frequency. The latter of these types was the most frequently thrown dur- 

 ing the early period of my work though it has been rare recently. On the 

 other hand, the former type, which rarely appeared at first, is now the most 

 common. A type characterized by thickness of conidium, as M6, M8, etc., 

 has been frequent all the time. These three types were by far the most 

 common, and may be said to show the three tendencies. Markedly short- 

 conidia saltants were few, as were also clump-bearing types that possessed 

 permanence. Strains that threw either of the two types first mentioned 

 above were very likely to continue to throw similar types. The same may 

 be said of clump-bearing types. 



STABILITY OF THE SALTANTS 



Many saltants have been tested in various ways to determine, to some 

 degree, their constancy. Through numerous transfers on corn-meal agar 

 the O colony and the M colony of many saltants have been carried side by 

 side. Under such conditions, though the original may give rise to new 

 saltations or the saltant may saltate further, the main portion of both the 

 O and M colony, as a rule, maintains its characters. 



It is manifestly impossible to test all the saltants to ascertain what 

 their future behavior will be. All that can be done at present is to record 

 certain observations concerning them. Several saltants possessing strongly 

 distinctive characters have been repeatedly transferred and have maintained 

 their characters through all of these transfers ; and as far as can be foreseen 



