INTRODUCTION 9 



scolecospore, and the most satisfactory solution has been found to lie in the 

 ratio between length and width. A compilation of all the long-spored species 

 of the one and short-spored of the other in Sphaeriales discloses the fact 

 that a ratio of 20:1 represents much the most natural dividing line and one 

 that requires the transfer of very few species to make it consistent. A simi- 

 lar study of the Phomales demonstrates that a ratio of 10:1 is preferable, 

 the difference being probably explained by the normally smaller size of the 

 pycnidium. 



It is obvious that the determination of spore characters must rest upon 

 mature spores; this is especially important in Ascomycetes where maturity 

 is sometimes long delayed, winter conditions apparently being often neces- 

 sary to insure this in nature. Spore color andseptation, as given in descrip- 

 tions, are frequently misleading or erroneous, as color and septation usually 

 depend upon age and condition of development of the spores. Spores in 

 some cases, e.g., Macrophoma, may be discharged and appear mature and 

 germinate freely, while later the spores remaining in the pycnidium become 

 brown as in Sphaeropsis and sometimes septate as in Diplodia. This is also 

 true of spore septation, which in some cases is delayed until after the spores 

 seem to be fully formed and mature and are expelled. Only careful observa- 

 tion of abundant material in different stages of development can determine 

 these points in any particular genus. An examination of the older type 

 specimens by Hoehnel and others has shown that the original descriptions 

 were sometimes based upon immature material that failed to indicate the 

 true nature of the spores as to color and septation, but this in no wise 

 detracts from the value or usefulness of these characters under the proper 

 safeguards. As with all fungus criteria, much more careful observation, 

 statistical study and experiment are necessary to determine just how stable 

 and dependable these characters are in representative genera, as well as in 

 spore sections and higher groups. 



Stroma 



Among the Sphaeriales and Phomales in particular, no other structure 

 is so variable and so difficult of interpretation as the stroma, probably 

 because this is a part of the vegetative body and hence more directly affected 

 by the environment. Perhaps the major part of the disagreement between 

 Hoehnel, Theissen and Sydow, and Petrak centers about the facts as to the 

 various types of stromata and their interpretation. The rise and fall of the 

 Pscudosphaeriaceae is the outstanding illustration of this, but it is likewise 

 exemplified throughout the stromate and stromoid forms. Probably more 

 new genera have been proposed on the basis of differing interpretations of 

 the stroma than on any other. Petrak in particular has recognized the unde^ 

 sirability of this, but has not always heeded his own conclusions (Ann. Myc. 

 21 : 272 1923; 23 :83 1925). It is obviously true, as Winter pointed out in 

 discussing Sordaria and Hypocopra, that many mistakes have been made by 

 basing genera on stromatic characters alone (1887:169). 



