THE PERITHECIUM OF MELIOLA. 17 



our Erysiphece, and Berkeley 1 takes the same position. These opinions 

 appear to have been based simply on the resemblance in habit and 

 the more obvious anatomical characters, and on the fact that no Ery- 

 siphe is known in the tropics. 



The details of structure, and especially of the development of the 

 fruit-bodies above described, enable us to criticise these views from a 

 somewhat firmer standpoint. 



Apart from minor points of resemblance between Meliola and the 

 typical Erysiphece, such as the haustoria (not well developed in 

 Meliola), the asci, &c, there can be no question as to certain points of 

 agreement in the structure and development of the perithecia; 

 nevertheless, the origin of the fruit-body in the two groups is not 

 obviously similar, and at first sight the differences may seem greater 

 than they really are. 



In the typical simpler Erysiphece, such as Podosphaera, as is well 

 known from De Bary's classical researches, 2 the carpogonium and anthe- 

 riclium arise each as a short lateral branch from separate hyphse, at 

 the point where two hyphee cross : each becomes cut off by a septum, 

 which is formed close to the parent hypha in the case of the pyriform 

 carpogonium, and about half way up the curved antheridium branch. 

 The free end of the latter becomes closely applied to the top of the 

 carpogonium, and fertilisation — possibly not complete in a physio- 

 logical sense, however— is said to be complete. After this process 

 numerous branchlets arise from the base of the antheridium filament 

 (and also from the base of the carpogonium), grow rapidly and with 

 numerous segments, and invest the carpogonium, which meanwhile 

 begins to be (more slowly) cut up into cells. 



In Eurotium s we have an essentially similar process, escept in 

 minute details, and the antheridium is a branch springing from the 

 same hypha which bears the carpogonium, and arises just beneath the 

 latter. Here, as before, the perithecium envelope is formed chiefly by 

 the rapid overgrowth of cells derived from the antheridium branch. 

 It is quite conceivable that a form allied to Erysiphe and Eurotium, <kc, 

 might have the unicellular carpogonium and antheridium arising quite 

 in contact at their bases from the same branch. 



If we now compare the above with the succession of events in the 



1 Introd. to ' Crypt. Bot.,' p. 275. 



2 'Beitr. z. Morph. u. Phys. d. Pilze,' R. iii., 1870 

 8 C/., Pe Bary, loo. cit. 



