Pound and Clements: north American hyphomycetes. 645 



forms, and even the conidial forms of the same genus 

 differ so greatly as to be placed in the most diverse 

 groups when found only by themselves. It is practically 

 impossible to arrange the form-genera of the Hyphomycetes 

 according to the mature forms. A considerable number are 

 now known in all stages. For many species the perfect forms 

 maybe guessed at with more or less certainty. For more, they 

 cannot be stated at all. In the case of a large part of the re- 

 ported species the perfect forms are wholly unknown, if there 

 are any. Such an arrangement being out of the question, the 

 only plan would seem to be to arrange the forms according to 

 their apparent structural relationship, as we would do in an 

 autonomous group. This we have attempted to do. When we 

 speak of relationship we must be understood as referring solely 

 to structural similarity. In the case of autonomous forms this 

 would be taken as an index to genetic relationship. In such a 

 group as the Hyphomycetes it can only be used as a basis of 

 arrangement in dealing with forms with which collectors must 

 constantly have to do. 



It may be asked whether in such a group as the Imperfect 

 Fungi, an artificial arrangement after the manner of a key, 

 such as that adopted by Saccardo, is not the best. In a certain 

 sense there can be no natural arrangement of such a group. 

 But on the other hand a large portion of the Hyphomycetes have 

 a certain autonomy. They are constant, and they develop and 

 reproduce themselves indefinitely without attaining any differ- 

 ent form. Besides, the Saccardian arrangement is by no means 

 as easy of application as it may appear. Phragmosporous 

 genera with species having continuous conidia are not uncom- 

 mon, the line between Didymosporeae and Phragmosporeae is 

 always shadowy, and Dictyosporeae are very apt not to have 

 uniformly muriform spores. Experience with this arrange- 

 ment soon convinces one that unless he knows a genus, the 

 neatly planned system will often do little towards enabling him 

 to identify it. 



For such reasons we have determined to attempt a re-arrange- 

 ment of the genera based on structural similarity and relation- 

 ship, endeavoring to treat the form -genera in the same manner 

 in which one would deal with autonomous groups. 



In the Saccardian arrangement the Hyphomycetes are divided 

 into four families, the Mucedineae in which the sporophores are 

 free and always hyaline or light-colored, the Dematieae in 

 which the sporophores are free and hyphae and conidia are 



