Classification. 79 



CHAPTER XX. 



CLASSIFICATION, WITH SPECIAL REFERENCE TO BIOLOGIC FORMS. 



Our views as to the limits of species have undergone a change owing to 

 the results of infection experiments in the rusts as well as in other groups 

 of fungi. Species have hitherto been distinguished on morphological 

 grounds, those possessing the same structural characters being considered 

 identical, and separated from those which differ from them in essential 

 points. But in recent times, when infection experiments have been carried 

 out on an extensive scale, it has been found that parasitic fungi, completely 

 agreeing in structural characters, or at least differing so slightly as to be 

 incapable of separation, have very different infective powers. It has 

 therefore become necessary to recognise such forms, and since the dif- 

 ferences are based upon physiological or biological characters, they will be 

 distinguished as "biologic forms." Various names have been proposed 

 for these different varieties, such as "special forms," by Eriksson; "bio- 

 logical species," by Rostrup ; "sister species," by Schroeter ; and 

 "adapted races," by Magnus. 



The truly morphological species, such as Puccinia graminis, Pers. or 

 Erysiphe graminis, DC., have still to be recognised; but each one may be 

 split up into a number of different forms, with distinct powers of infec- 

 tion. 



GRADATIONS OF SPECIFIC VARIATION. 



There is every possible gradation, however, between species which are 

 morphologically distinct and those which can only be separated on biologic 

 grounds. Puccinia graminis Pers., for instance, is recognised as an inde- 

 pendent species, because all the forms of this rust produce aecidia on the 

 barberry (except where it has lost this power), and the structural charac- 

 ters are always practically the same. It is regarded by Eriksson as a 

 collective species, in which the different members are so closely related, both 

 morphologically and biologically, that they are only separable into biologic 

 forms, and not to be distinguished as species. The forms on the different 

 hosts are not identical, but they constitute a series, each member of which 

 runs its course on definite host-plants, and is more or less strictly confined 

 to them. Arranged according to the principal host-plants they are ^as 

 follows: i, secalis; 2, avenae; 3, airae; 4, agrostidis; 5, poae ; 6, tntici. 



The collective species known as Puccinia rubigo-vera (DC.) Wint. was 

 first divided in 1894 by Eriksson ,and Kenning 1 into the two distinct species 

 of P. glumarum (Schum.) Eriks., the Yellow Rust, and P. dispersa, Eriks. 

 and Henn. the Brown rust. P. glumarum has not been found in Australia, 

 and it is not necessary here to refer to the various biologic forms into 

 which it has been divided. No aecidial stage has been found in connexion 

 with it. 



P. dispersa, when first separated from P. glumarum, was split up into 

 four biologic forms by Eriksson, since he did not at that time consider 

 them sufficiently distinct to be designated species. These forms were^:- 

 i, secalis; 2, tritici ; 3, bromi ; 4, agropyri. Further investigation, 

 however, led him in 1899 to raise the various biologic forms to the rank < 

 independent species, and it will be interesting from our present stand- 

 point to consider the reasons given by Eriksson for differentiating between 



