72 PABASITES BELONGING TO THE GENUS GLOMEKELLA. 



unless some other basis than morphological differences and differ- 

 ences in behavior in culture media can be established as a basis for 

 species segregation. It should be shown either that there are morpho- 

 logical differences between the conidial form of the Fusarium causing 

 the wilt and the conidial form of Neocosmospora sufficient to identify 

 and separate them, or else a large series of inoculation experiments 

 should be carried out, using ascospores and conidia from a variety 

 of sources and including a sufficient number of s trams to justify the 

 conclusion that the fungus will not infect the host and may be con- 

 sidered a physiological species. 



The other suggestion of the same authors is that the species of 

 Fusarium with which they were working and which did not produce 

 perithecia in cultures maybe autonomous forms which have no perithe- 

 cial stages and that their close resemblance to the conidial forms of 

 Nectria, Neocosmospora, and related genera may be a mere super- 

 ficial one which does not necessarily imply any direct relationship. 

 This is ingenious also, but unfortunately it is incapable of experi- 

 mental proof and is not in harmony with the present trend of myco- 

 logical facts and opinions. 



The connection between various conidial and pycnidial forms and 

 their ascogenous stages is being slowly but surely proven by pure- 

 culture methods, and it seems much more natural and reasonable to 

 suppose that the majority at least of the so-called " imperfect fungi" 

 are stages in the life history of ascomycetes and hymenomycetes, 

 though comparatively few have yet been positively connected and 

 the determining factors concerned in the complete development of 

 the organisms are still unknown. 



The production of perithecia in the numerous cases of Glomerella 

 which are reported here is evidently a fairly well-fixed hereditary 

 racial character. Having obtained by repeated cultures from differ- 

 ent collections a race or strain which produces both conidia and peri- 

 thecia on culture media, other generations grown from either the 

 conidia or ascospores of this strain continue to produce both spore 

 forms of the fungus indefinitely. In one case, Glomerella cingulata 

 from Persea, as already described, a race was grown through 23 suc- 

 cessive generations and still produced both conidia and ascospores 

 in as great abundance as at first. 



Miss Wakefield (93), as a result of her investigations of Scliizo- 

 pliyllum commune and Stereum purpureum, finds also certain races 

 which show a much greater tendency to produce spores than others 

 under the same conditions of culture. Brefeld (14) states that 

 Penicillium produces ascogenous fructifications at one time while at 

 another, under exactly the same conditions, it produces only conidia. 



All the writers' work with Glomerella, as well as with other pyreno- 

 mycetes, indicates that the production of perithecia is a hereditary 



252 



