INTRODUCTION 7 



led to the duplication of genera in many cases and has little or no dependable 

 value except in special instances. 



At present, the use of habit as a generic criterion is firmly entrenched 

 in mycological practice, but it should be clearly understood that such charac- 

 ters while utilized in the Key are not necessarily considered of generic value 

 by themselves. Sufficient evidence is already available to show that such 

 criteria are in certain groups of little real worth and should be used with 

 great caution. Habit as a criterion appears to fail almost completely in the 

 Hypocrcaccac, where twenty of the larger genera occur on from three to ten 

 different types of host or matrix. 



CoNiDiAL Stages 



With increasing knowledge of the life-histories of the Ascomyceies, 

 conidial or "nebenfrucht" characters are being adopted in defining and 

 limiting old as well as new genera. Where sufficiently exact knowledge of 

 the development of the various species is available, this may ultimately prove 

 desirable, but too little information of this kind has been published to permit 

 any general application of such criteria in a key. Moreover, our present 

 scanty knowledge of the subject furnishes various examples of the difficul- 

 ties that arise in attempting to utilize conidial stages for generic segregation. 

 It has been found that ascogenous forms generally regarded as congeneric 

 have very different secondary stages, while widely separated genera may 

 possess similar or nearly identical ones. Further discussion of this theme 

 may be found in "The Problem of a Natural Classification of the 

 Asconiycctes" (Shear, 1929). Furthermore, some workers have gone so 

 far as to segregate genera on the basis of the mere association of certain 

 conidial forms with the ascocarp. Such practice is to be deplored, as it 

 can only lead to greater uncertainty and confusion. 



The names of the so-called form genera of Fungi Imperfecti, which in 

 most cases represent stages in the life-histories of Ascomycetes, should be 

 recognized as tentative, until their genetic relation to the perfect form is 

 definitely shown, when they can be reduced to synonomy and discarded, as 

 has already been done in the Pucciniales. For present purposes therefore, 

 the most convenient and usable artificial system constitutes the most desir- 

 able arrangement of this group. Such attempts as those of Hoehnel to 

 establish a new system of Fungi Imperfecti hence serve no useful purpose, 

 except in so far as they increase the readiness with which specimens in hand 

 may be identified. Whoever tries to use Hoehnel's key in this connection is 

 practically certain to concur in the judgment of Petrak, already quoted, that 

 it is much less satisfactory than the Saccardian. 



Spore 



The opinion is frequently expressed that the carpologic system of 

 Saccardo is much less natural than one based upon stroma and perithecium 

 as primary criteria. With our present knowledge, no objective determination 



