477] MELIOLA IN PORTO RICO — STEVENS 9 



influenced by the nature of the leaf upon which it grows, that it will 

 be more nearly straight on a straight-veined, smooth leaf than upon a 

 crooked-veined, hairy one. 



The question of limits of variation in mycelium, hyphopodia and 

 setae is a difficult one and finally will be solved only by inoculation 

 experiments and observations upon live material. 



I have left out of all consideration the so-called conidial stage, Hel- 

 minthosporium, Podosporium, etc., because these structures appear use- 

 less in taxonomy and, moreover, because I am far from convinced that 

 they really are conidial stages of the Meliolas. The mucronate hyphopodia 

 are curious structures but they are so variable on the same mycelium 

 that they possess little value in classification. 



The question of biologic specialization has been but little studied 

 in this genus. My field observation, however, strongly supports the idea 

 that there is large specialization. For example, Cissus vines heavily 

 coated with M. merrillii are found with their branches interlacing 

 with plants susceptible to Meliola but the Cissus Meliola does not invade 

 the other plants. The Meliolas on Dieffenbachia and Philodendron, mem- 

 bers of the same family, to the naked eye appear mu^li alike and under 

 the microscope are kin, yet there is no field evidence that one goes to 

 the other's host. In some cases forms morphologically indistinguishable 

 are found upon large numbers of species or genera in the same family 

 (notably 31. hicornis on the Leguminosae ) , whether these forms are inter- 

 inoculable or not is not known. 



It is possible that were type specimens of all the old species at hand 

 for comparison some of the new species might be found to agree with 

 them, but from the descriptions and figures available such does not seem 

 to be the case. It is more probable that some of the collections which 

 I have referred to old species, relying upon printed descriptions and 

 without comparison with types, in reality may be new species. It seems 

 best for present purposes, however, to avoid undue multiplication of 

 species and I believe that my error lies more largely in conservatism than 

 in the creation of new species. 



It is extremely difficult to interpret many of the descriptions that 

 have been written, some are too brief and leave out many essentials, 

 others describe setae but do not state whether they are perithecial or 

 mycelial and the supposition is strong that there have been many cases 

 where the colony character has been taken from a colony parasitized by 

 some other fungus, e.g., many of the Meliolas described as having a 

 crustose colony are probably Meliolas parasitized by a Microthyriaceous 

 fungus. 



The final word as to the specific relation cannot be said but with 

 the large number of collections at my disposal, on many hosts, and the 



