10 GENERA OF FUNGI 



With regard to the dependabihty of such criteria, this may vary greatly 

 in different families and even in genera. In some the stroma is fairly con- 

 stant, in others very variable; effuse, valsoid and pulvinate forms often 

 intergrade, as do sessile and stipitate ones also. However, the most confus- 

 ing cases are those connecting Sphacrialcs and Dotliidcalcs, in which the 

 question arises as to whether a stroma contains perithecia or locules. Here 

 again all possible intermediates occur between stromata in which the peri- 

 thecia are so distinct that they are easily removable, to those in which the 

 asci are borne in a chamber of the stroma which shows no definite wall. 

 Hoehnel and others have gone so far as to recognize a separate family, 

 Pseudosphaeriaceae, to include genera such as Plcospora and Pyrenophora, 

 in which the perithecia have a somewhat thickened wall that they regard as a 

 stroma with a single locule ! In this connection, it is to be noted, that Blain 

 has found that stromata "possessing interascicular pseudoparenchyma, the 

 distinguishing feature of the Psciidosphaeriales, are found in the Dotliidcalcs 

 and Sphacrialcs" (1927:18). 



Recently, Miller has attempted to distinguish between a perithecium 

 and a stroma with a single locule. He concludes that the perithecial wall 

 in the Sphacrialcs "is histologically and ontogenetically different from the 

 tissue of the stroma," and defines it "as the specialized tissue which arises 

 from the archicarp, and from the beginning encloses the ascigerous cen- 

 trum." It is also stated that the ostiole in a true perithecium is schizogenous 

 in origin, while in the locule it is lysigenous (1928:194). Whatever the 

 actual facts are in the case, in order to determine them conclusively and 

 make them available for practical taxonomic purposes, further investigation 

 embracing many more genera and species is imperative. 



Insertion 



The position of the ascocarp, stroma or pycnidium with reference to 

 the tissues of the host, i.e., whether innate or supei-ficial, has long been 

 regarded as a criterion of generic significance, and the distinction has been 

 applied with almost complete consistency to the orders concerned. There has 

 been some further tendency to distinguish erumpent forms, but these present 

 the double difficulty of discriminating between both normal insertions, quite 

 apart from the wide variation in the degree of erumpence itself. Innate 

 insertion is likewise modified by concretion with the epiderm to furnish an 

 additional generic criterion. 



However, Theissen and Sydow in the Dotliidcalcs (1915) and Hoehnel 

 in the Phacidialcs and the stromoid Fungi Iinpcrfccti have carried this dis- 

 tinction to extremes and have segregated a host of new genera with respect 

 to origin between cuticle and epiderm, between epiderm and mesophyll, or 

 within the latter. The difficulty of determining the facts in many cases and 

 their known invalidity in others prohibit for the present at least the use of 

 such criteria. Still more serious is the fact that the proponents disagree as 

 to the facts in a number of critical cases; for example, Hoehnel states that 



