I9c8] . EDGERTON—ANTH RACNOSES 389 



Sections of perithecia a little past maturity showed the spores held 

 together in a mass just outside and at the apex of the neck. The 

 whole inside of the perithecium seemed to be empty, with the excep- 

 tion of a few scattered threads. It seems probable from the sections, 

 although this was not determined absolutely, that the ascospores are 

 held together by a mucilaginous substance as well as the conidia. 



The question of paraphyses is a much discussed one at present. 

 The genus Glomerclla w^as described by Spaulding and von Schrenk 

 as without paraphyses, but quite recently Sheldon (39) has reported 

 their presence in a number of forms, especially in the young perithecia. 

 Shear (40) has taken exception to Sheldon's observations, finding 

 sterile threads in the perithecia but believing they were entirely out- 

 side of the asci, so they would be periphyses and not paraphyses; 

 later (41J, however, he calls them evanescent paraphyses. The 

 point is an extremely difficult one to decide. Forms from several 

 hosts have been studied to determine this point. In crushing a fresh 

 perithecium, sterile threads can be easily seen; jig. 42^ drawn from 

 the form on Dracaena, shows them plainly. These threads are very 

 irregular, many of them much longer than the asci, sometimes branched 

 and sometimes septate. They are very highly granular, in fact the 

 granules are the most prominent part, the wall itself being almost 

 invisible. The difficult question is to determine whether they are 

 between the asci or outside of them. Scores of stained sections have 

 been prepared and carefully examined; some slides have looked a 

 little suspicious, yet never have the threads been seen between the 

 asci with certainty. While the threads take the stain poorly, they can 

 be seen, and appear to form a layer around the asci near the base of 

 the perithecium, but in- the upper part of the perithecium above 

 the asci they branch out over them. As the asci and spores break 

 loose from the bottom of the perithecium, they can be seen mixed in 

 with threads. They could hardly be considered evanescent para- 

 physes, because they remain in the perithecia as long or longer than 

 the asci. It seems best, therefore, to regard the genus as apara- 

 physate. 



The perithecial stages from all of the forms, so far as known, seem 

 to be quite similar. The most variable characters are the size of the 

 spores and the length of the neck of the perithecium. The spores 



