Freeman: SYMBIOSIS IN THE GENUS LOLIUM. 331 



This experiment also failed and a consideration of the nature 

 of the fungus might lead one to predict this result. Very 

 closely adapted parasites thrive best on healthy plants, so that 

 during the vegetative period conditions favorable to the host 

 would favor the parasite also. Whatever the true nature of the 

 Lolium fungus or its ancestors may be, there is no doubt that 

 it is now a very highly specialized parasite. Consequently the 

 weakening of the host did not favor a tendency toward spore 

 formation, in the predominance of parasite over host. These 

 experiments therefore only strengthen the supposition that the 

 spore forming power of the fungus has disappeared entirely. 

 Further cultural investigations have confirmed the result of 

 previous anatomical research on the life-history of the fungus. 

 Twelve plants of L. temulentum with the fungus were planted 

 in pots of three each at the Cambridge Botanical Gardens in 

 Cambridge, Eng., and six were left in the open while six were 

 placed under glass bell jars during the flowering season. This 

 precaution was taken to guard against the possibility of the 

 transferrence of fungus spores with the pollen. It should be 

 mentioned here however that no fungus hyphag have been found 

 in any part of the stamens except at the base as in other leaves. 

 Twelve plants without the fungus were placed under similar 

 conditions. The entire crops of these plants were sent to me 

 in the fall of the same year. The number of grains received 

 in each case was as follows : 



From plants with the fungus in the open 3?59^ 



From plants without the fungus in the open 222 



From plants with the fungus covered at flowering 1*071 



From plants without the fungus covered at flowering 824 



Out of these 100 of each were examined to determine the 

 presence of the fungus. In every case 100 per cent, came 

 true to the parent plant. Of another lot of with-fungus plants 

 100 grains were examined and again 100 per cent, were found 

 infected. This should establish beyond a reasonable doubt 

 both the accuracy of this method of infection and the existence 

 of two- races of L. temulentum one with and the other without 

 a fungus and further supports my previous researches on the 

 modus operandi of infection. 



It is a very noticeable fact that the crops of these Cambridge 

 plants show a considerable difference in the number of grains 

 produced by the two races and that the plants with the fungus 



