MICHIGAN ACADEMY OF SCIENCE. 9«7 



in the field whose leaves were badly infected with tlie disease were often 

 examined but the fungus was never found growing upon them. 



It was found tliat the fungus grew readily upon disinfected seed 

 (soaked in HgCL. 1:1000 for 20 minutes, and rinsed in sterile distilled 

 water) which had just enough water present to allow the seed to germi- 

 nate. The growth of the fungus was rapid and abundant. Within two 

 weeks the entire mass of seed was well overrun, and an abundance of 

 spores produced. The seed coats were almost completely covered by the 

 fungus and eventually destroyed. The cotyledons of germinating seeds 

 could be seen emerging from a black fungus shell of what had been the 

 seed coat. The cotyledons were not attacked, though some of them were 

 slightly discolored, but the tip of the radical, which in some cases attained 

 the length of 10 mm. was attacked by the fungus and killed, thereby 

 preventing further growth. It is possible that as the seed dried out some- 

 what, the growing tip died and was attacked saprophytically. However, 

 some of the radicals were attacked just as they were emerging from the 

 seed. The large bulk of the seed was so completely enveloped by the 

 fungus that they could not even germinate. Some of these seeds were 

 planted but failed to grow. 



From these experiments and from the observations made by Milburn 

 already referred to, there is a possibility of seed carrying the fungus as 

 threads within tlie seed coat. Tlie failure of attempts to infect the grow- 

 ing flower parts, and the absence of characteristic lesions on the young 

 cotyledons seem to indicate that the fungus threads within the seed coat 

 are not especially important in causing infection of the young plant. The 

 discussion of this latter phase is taken up under the topic. Dissemination. 



Htimidity and Infection: 



In the series of infection experiments already reported, care Avas taken 

 to insure a humid atmosphere after inoculation. The germination studies 

 reported on page 304 proved that the lowest relative humidity at which 

 air dry spores will germinate lies between 92.8 and 93.4%. 



Exi^eriments with Septoria lycopersici performed in tliis laboratory 

 (Levin, 1916) has shown that this pathogene could infect without the 

 film of water commonly held essential for germination. It seemed, there- 

 fore, proper that the relation of infection by the clover fungus be 

 similarly determined. Red clover plants were inoculated by placing small 

 masses of air dry fungus growth (culture 10 daj^s old) upon the surface 

 of many of the leaflets. The plants were then well watered and the soil 

 and outside of the pots completely covered with paraffin to prevent the 

 evaporation of water from any source but the plant. These plants were 

 l^laced under a sealed bell jar, through which air, previously dried by 

 passing through HoSO^ and CaCL, was drawn by means of a suction 



