8o 



Classification. 



the biologic forms of P. dispersa and those of P. graminis. The name of 

 P. dispersa Eriks., was reserved for the form on rye, and it is separated 

 from that on wheat, for example, on the following grounds: 



1. The uredo is only able to infect the host-plant, Secale. 



2. This form regularly appears several weeks earlier than that on 

 wheat, even when the host-plants are growing alongside of each other. 



3. The aecidiospores on Anchnsa spp. only infect Secale, and the teleuto- 

 spores of this form can in turn only produce the aecidJa on Anchusa. 



4. The teleutospores are only capable of germination immediately after 

 they are formed, while those of' the others only germinate in the following 

 spring. 



It will be observed that these are only biological or physiological dif- 

 ferences? but in the two species recognised by Eriksson which occur in Aus- 

 tralia, viz., P. triticina and P. bromina, there are also structural differences 

 sufficient to separate them apart from other considerations. In P. bromina 

 the uredo-sori are much larger, and the bright orange uredospores are also 

 slightly larger. It is in the teleutospores, however, that the difference is 

 most striking, and those of P. triticina are much nai rower, only being 

 about three-fourths that of the other. 



P. hieradi (Schum.) Mart., as at first constituted, was a collective species, 

 but. like so many others, when the test of infective power is applied to 

 them, it had to be broken up into several. It has been proved by Jacky 1 

 that the common Puccinia occurring on species of Hieracium can only infect 

 this, and not other genera of Composites, so that the original name is 

 restricted to the species found on this genus. Other species of Puccinia 

 occur on Hieracium, but (they are readily distinguished by their different 

 morphological characters. Similar results have been obtained with Puccinias 

 occurring on> other genera, so that they also must be regarded as distinct 

 species. 



It will be observed that only biological characters are here taken into 

 account, depending on the nature of the host-plant, but it is believed that 

 the closer and more careful investigation now rendered necessary will result 

 in structural differences being found, possibly of a more minute and less 

 striking character than formerly, where they were hitherto unsuspected 

 and overlooked. 



The three types selected for illustration will show the spirit of our 

 classification. 



P. graminis is the type of a single species which is split up into a number 

 of biologic forms, agreeing in the faot of producing aecidia on the bar- 

 berry, and the aecidiospores in turn infect the grasses. 



P. dispersa, as originally constituted, is the type of a collective species 

 at first divided into biologic forms, which were afterwards found *to be 

 distinct species, both on morphologic and biologic grounds. And the 

 original P. hieradi is the type of a collective species, the members of which 

 Strictly limited in their infective power to single genera, and their 

 main title to distinction as species is this biological peculiarity. 



The phenomenon of specialisation, first discovered by Eriksson, has 

 hus caused us to revise our conceptions of species, and it will be necessary 



the future not only to include under that name such forms as possess 



:t morphological characters, but also those which, although morpho- 



similar, are yet confined to definite host-plants. It will thus be 



necessary to split up the old species of P. hieradi into several of equal rank, 



the one growing on Hieracium spp. will form a type round which the 



others may be grouped. Just as in the old species there were differences of 



opinion as to the amount of structural differentiation necessary to separate 



