286 S alm o n , On Specialization of Parasitism in the Erysiphaceae. 



On accoimt of the apparently contradictory nature of the evi- 

 dence obtaiiied, and on account of the small number of the experi- 

 ments and the lateness of the season in which they were carried 

 out, it would obviously be unsafe to attempt at present, by con- 

 clusions drawn from the above experiments , the Solution of the 

 Problem of the behaviour of the spores in the first generation of an 

 Oidium transferred to a new species of host-plant. 



There is every reason to believe that there exists within the 

 genus Bromus — in addition to the four, or moie probably five, 

 Oidia shown at Table 9, a considerable number of ,biologic forms.' 

 Erysiphe Graminis has been recorded (1) (2) in the perithecial stage 

 on the following species of Bromus; B. mollis, B. secalinus, in the 

 section Serrofaicus', B. madriiensis^ B. rubens^ and B. sterilis in 

 Stenobromus; B. asper and B. hreviaristatus in Festucoides; and J5. 

 unioloides in Ceratochloa. Further, the Oidium. of E. Graminis 

 occurred in the Cambridge Botanic Gardens on Bromus squarrosus 

 [Serrafalcus) and B. arduennensis [Liberfia). 



The fact may be mentioned here that all the Oidia, used in 

 the experiments, which occurred in quantity on plants of B. inter- 

 ruptus, B. /wrdeaceus, B. com?nutatus and B. racemosus died away 

 without producing any perithecia in the autunm. 



In discussing the distinctness of these ,biologic forms,' we have 

 thus far been leaving out of consideration the fact that the fungus 

 in question produces other reproductive bodies, viz. the sexually 

 produced ascospores. The whole problem becomes changed and 

 extended when we take into consideration the possible infection- 

 powers of the ascospores. Although nothing is at present known 

 by experiment on this subject, Neger has expressed the opinion 

 that it is probable that the ascospores serve as a connecting- 

 link between ,biologic forms.' Neger supposes that in the case 

 of two ,biologic forms' of Oidium growing on host-plants x and y 

 (where the Oidium on x is unable to infect y and vice versa) the 

 ascospores of the fungus produced on either plant will be found 

 capable of infecting the other. Neger cites as evidence in support 

 of this view certain phenomena he has observed in the case of some 

 species of Erysiphe on certain host-plants; and also adduces ana- 

 logies drawn from the üredineae. In the case of the phenomena 

 quoted it seems to me that the deductions drawn by Neger are 

 hardly Warrant ed. Neger observes (4, p. 270): Der Umstand aber, 

 dass auf einjährigen Pflanzen, z. B. Senecio vulgaris (bei welchem 

 also Mycel-Überwmterung ausgeschlossen ist) , eine Erysiphe sich in 

 jedem Jahr reichlich entwickelt, ohne indessen je zur Perithecien- 

 bildung zu gelangen, legt die Vermutung nahe, dass mittelst der 

 Ascosporen die Übertragung eines Mehltaupilzes von einer krt auf 

 eine andere (Wirtspflanze) wohl möglich ist. Demnach wären 

 die Ascosporen dadurch ausgezeichnet, dass sie das 

 Bestreben zeigen, den Kreis der Wirtpflanzen eines 

 Pilzes weit zu erhalten, während die Konidien sich sehr 

 schnell einem bestimmten Substrat anpassen.' Neger 

 also mentions (1. c. p. 258) the similar behaviom^ of the conidial 

 condition of an Erysiphe on Calamintha acinos, Symphytum tuberosum, 



