THE PEODUCTIOK OF PEBITHECIA IN GLOMEBELLA. 71 



nonproduction of perithecial forms in cultures. One is that the forms 

 which produce perithecia in cultures on artificial culture media are 

 pure saprophytes, whereas the forms which do not produce perithecia 

 in cultures on artificial culture media are parasites which require the 

 living host or living plant tissues to complete their development. 

 This scarcely accords with the well-established fact that there are all 

 grades and degrees of parasitism among fungi, from the highly 

 specialized obligate forms, such as rusts and smuts, to the weak faculta- 

 tive forms which are only parasitic under certain conditions and in 

 certain stages of their development and may be saprophytic under 

 other conditions or in other stages. A large part of the pyrenomy- 

 cetes, though parasites, are not of a high type. They frequently 

 pass the early stages of their development — that is, their conidial 

 and pycnidial conditions — as parasites on the living host, whereas 

 they only complete their life history and produce their perithecial or 

 sexual forms upon the dead tissues of the host after it has been 

 killed either by the direct action of the fungus or by some other 

 cause. This fact is emphasized by Massee (59) in his discussion of 

 the evolution of parasitism in fungi. It would not, therefore, appear 

 remarkable if many of these forms should complete their life cycles 

 on sterile culture media containing the essential nutrient substances 

 required by the organism. There is apparently no reason to suppose 

 that the case of Fusarium and Nectria is not directly comparable in 

 this respect with that of Gloeosporium and Glomerella. 



There seems to be an abundance of proof to establish the parasitic 

 character of many forms of Gloeosporium. In case of the bitter-rot 

 of the apple, for instance, this parasite has been obtained directly 

 from the tissue of growing fruits before their maturity and when 

 transplanted on artificial culture media has produced perithecia as 

 well as conidia. Perithecia are also frequently produced upon 

 fruits which have been destroyed by the parasite. The suggestion, 

 therefore, that in cases where a Fusarium does not produce perithecia 

 in culture media it is a parasite and that when a Fusarium does 

 produce perithecia in culture media it is a saprophyte, does not 

 appear to accord with what is already known in regard to Glomerella 

 and some other pyrenomycetes. 



The case of Neocosmospora, as described by Butler (16), is scarcely 

 more conclusive. The fact that he did not succceed in infecting 

 plants by the use of ascospores of Neocosmospora in the few attempts 

 made does not necessarily prove that this fungus is not a parasite 

 and that it is specifically distinct from the Fusarium form with which 

 he made successful inoculations. As is well known to plant and 

 animal pathologists, there is a great range of variability in the virility 

 or capacity for infection of various races, strains, or forms of what 

 are regarded as the same species; at least they must be so regarded 



252 



