NATURAL AND ARTIFICIAL INFECTION. 51 



The exact proof, however, is best obtained by means of experi- 

 mental infection. 



With many parasites the sporocarps are normally developed 

 saprophytically on a dead substratum, so that if parasitism 

 be suspected it can only be proved by infection. Thus the 

 perithecia of Ncctria cinnaharina develop only after the death 

 of the plant-organ, which the fungus attacked when alive. The 

 more complex reproductive organs of many fungi are developed 

 only on dead remains of the host, while on living or dying 

 parts one finds various forms of conidia of doubtful relation- 

 ship. In many cases it has been possible, by means of artificial 

 culture alone, or combined with artificial infection, to prove 

 various forms of reproductive organs to be stages in the life 

 of the same fungus. 



When a group of fungi contains both saprophytes and parasites, 

 it is often necessary to determine whether some species is para- 

 sitic or purely saprophytic. This is particularly the case with the 

 groups of Pyrenomycetes, Discomycetes, Hymenomycetes, several 

 groups of the lower Fungi, the Bacteria, and Myxomycetes. It 

 is unnecessary, however, with the Uredineae, Ustilagineae, Per- 

 onosporeae, Exoasceae, and other groups known to contain 

 parasites exclusively. 



Put even in these last-mentioned groups experimental in- 

 fection is necessary for obtaining information on other points. 

 The reproductive organs of Uredineae cannot be reared in 

 artificial solutions, so that their cultivation must be carried out 

 on the living host-plant. In this way alone can we ascertain 

 the relationship of uredospores, teleutospores, and aecidial-forms, 

 where any doubt occurs as to their belonging to the same 

 species. Infection becomes particularly valuable when one has 

 to investigate heteroecious Uredineae, whose various forms of 

 reproductive organs inhabit several host-plants. Thus it was 

 by means of infection that De Bary discovered the connection of 

 Aecidiiim berhcridis on the barberry, and Pitccinia graminis on 

 cereals ; likewise Hartig, the relationship of Melampsora Goep- 

 pertiana on cowberry with Aecidimn columnare on needles of 

 silver fir. There still remain many aecidia, teleutospores, and 

 uredospores, whose related forms have not yet been found. 



Infections are also necessary to determine the species of a 

 fungus. It has been found, for example, that Gymnosporanfjiiim 



