66 PAKASITES BELONGING TO THE GENUS GLOMERELLA. 



ASCOSPORES. 



The ascospores show almost as great variation in size, shape, and 

 other characteristics as do the conidia. They range from 9 to 24 by 

 3 to 7.5 p. The contents are sometimes rather coarse and granular, 

 at other times more homogeneous. Vacuoles or oil drops are fre- 

 quently present, sometimes one large one in the center or at each 

 end, at other times two smaller ones at each end. Ascospores are 

 almost always slightly curved. They are usually almost colorless, 

 but when old and in mass they frequently show a pale-lemon or 

 salmon color, and occasionally very old spores are decidedly dark 

 colored. 



PARAPHYSES. 



The early investigators of Glomerella do not mention the presence 

 of paraphyses. It will be seen by consulting Table II that these 

 bodies have not been very frequently found in our material or reported 

 by others. When observed they appear to be very thin walled and 

 are not always easily discerned. Their presence or absence does not 

 seem to be sufficiently constant to be of much taxonomic value. 



HOST RELATIONS. 



Most of the forms from various hosts appear to be able to infect 



other hosts under certain conditions, at least, as indicated in Table 



IV, which gives the results of cross-inoculation experiments of other 



authors as well as of the writers. With a few exceptions, such as 



the bean and cotton anthracnose and perhaps the form from the 



squash, different races from the same host seem to vary as greatly in 



virility as do forms from different hosts. 



lac 

 CAUSES OF VARIATIONS. 



In dealing with the phenomena of heredity and variatior.in Glom- 

 erella there is apparently no reason to believe that the Mendelian 

 theory is involved or that heterozygosis takes place, as no union of 

 nuclei between different individuals or species is known to occur. 

 Edgerton (32) has recently expressed his belief in a cross-fertilization 

 between two strains of Glomerella from Populus, but the evidence 

 given is not conclusive. 



It seems clear that whatever fusions may occur between nuclei in 

 the development of the individual organism which arises from a single 

 spore, such fusions, though they be admitted to repres3nt sexual 

 union, could scarcely be conceived to add or transmit any new char- 

 acters to the resulting progeny unless they were characters acquired 

 by one of the fusing nuclei and not by the other during the life history 

 of the individual. It is difficult to conceive how the union of two 



252 



