392 BOTANICAL GAZETTE [jttne 



sent one species, or at best varieties of one species. They base 

 their beHef chiefly on the fact that the perfect stages are not morpho- 

 logically distinct; but because two forms have similar perfect stages 

 it does not necessarily follow that they are the same species. In 

 higher plants there are many instances of the flowers of different species 

 being very similar, if not identical, yet no one would hesitate to call 

 them distinct; differences in the vegetative stage, as the shape or 

 lobing of the leaves, may be marked and constant and hence suffi- 

 cient to separate them in distinct species. The sexual stage is the 

 least subject to modification, and we must expect very closely allied 

 species to be very similar in this stage. This must be especially 

 true in forms in which the perfect stage is only occasionally produced, 

 as seems to be the case with Gloeosporium. 



Theoretically, if we should grow two forms side by side on the 

 same medium and should watch their development, we ought to be 

 able to tell whether or not they are distinct. If the growth characters 

 were different, if the colonies formed were distinct, or the effect on 

 the medium were not the same, we should doubt tho advisability of 

 calling them the same thing. Miss Stoneman carried several f'^rms 

 through oa different media and was able to separate some of the 

 different forms by means of these characters. Thi'? method has 

 long been in use in bacteriology for det^-rmining species; and recently 

 the method has been advocated by Thom (45) for determining species 

 of Penicillium. He has used it in his work on this genus and has 

 found it satisfactory, claiming that the characters of a species on a 

 standard medium are alwaj^s the same. No matter how long a species 

 may be grown on some other medium, when it is brought back to the 

 standard medium it immediately returns to its original characters. 



A thorough attempt was made to test this method with the dif- 

 ferent members of the GlomereUa type; all of these have not developed 

 the ascigeral stage, but they undoubtedly belong t(^ this type. Cul- 

 tures from over twenty different hosts were used. Petri dishes, 



with a certain medium 



pure culture. 



from ail the similar 



same time, generaUy three plates of each form. The medium for 

 the set of cultures was made in one batch, so that there was no chance 

 for variation in this factor. Many sets of cultures were made cm- 



