Freeman: SYMBIOSIS IN THE GENUS LOLIUM. 333 



pushed further forward it must necessarily take place in the 

 intraseminal life of the host. Granting this to be the case, one 

 would expect to find the infection finally taking place as soon 

 as the stem growing point is established, for not until then are 

 the chemotactic substances present to attract the hyphge. And 

 this is what actually occurs. 



An analogous driving forward of the period of infection has 

 probably taken place among those parasites attacking the host 

 plant in the extraseminal seedling stages. It seems not improba- 

 ble that the oat smut may have formerly been able to infect any 

 immature tissues of the host as the corn smut does. The ovary 

 is a particularly favorable organ both on account of nutrition 

 and distribution facilities and hence, the preference for the 

 ovary as a spore-bearing region for the parasite. The estab- 

 lishment of an attraction by the growing point of the stem and 

 the localization of the sporiferous region in the ovary would 

 naturally be followed by an earlier infection since the presence 

 of the fungus in the unbranched stem multiplies the results of 

 the infection when the stools are mature. The proximity of 

 spores to the grains in sowing becomes obviously advantageous 

 and hence the infection would be pushed forward to the earliest 

 stages of the extraseminal development of the host, as in the 

 oat smut. If now, one presupposes a chemotactic attraction of 

 the growing point for the hyphae of a smut and also at the same 

 time a failure of the fungus to destroy the endosperm and young 

 embryo during spore formation, then nucellar hyphae crowded 

 back by the endosperm and embryo, might easily effect direct 

 infection of the growing point of the young embryo. This 

 period would then be shoved forward to the formation of the 

 earliest rudiments of the growing point. It seems probable that 

 the fungus is therefore an ustilagine which has brought forward 

 its infection period to the intraseminal stages of the host and 

 more particularly to the first appearance of the growing point, 

 and has lost the power of spore formation. 



Apparently infection of a without-fungus plant of L. temulen- 

 turn is^impossible as is also the eliminating of the fungus from 

 the with-fungus plants, barring the possibility of failure to in- 

 fect all ovaries in an inflorescence, which case seems highly 

 improbable. There exists therefore not only two races of L. 

 temutentum, but also of L. -perenne and of L. linicola and 

 probably of other species of Lolium, in each case one with, and 

 the other without, the fungus symbiont. 



