

NATURAL HISTORY SOCIETY OF GLASGOW. 299 



P. Poarum, Kiel. iEcid. on Tussilago Farfara. Tel. on Poa annua and 

 P. pratensis. 



P. Magnusiana, Korn. jEcid on Rumcx Eydrolapaihum. Tel. on Pkrag- 

 miles communis. 



P. sessilis, Schneid. iEcid. on Allium ursinum. Tel. on Pkalaris arundi- 

 nacea. 



P. caricis, Sehuni. JEcid. on Urtica dioica. Tel. on various carices. 



P. sijlvatica, Schrbt. iEcid. on Taraxacum officinale. Tel. on Carex 

 muricata. 



Assuming that the main facts of this plant heteroecism are 

 correct, it is probable that only a few of the connections and 

 relations have been traced, and that, as our knowledge grows, inter- 

 communications at present undreamt of may be discovered. For 

 instance, JEcidium Grossularice, which is ranked under section 

 Pucciniopsis, is probably heteroecism al. Though common in this 

 country, yet here no teleutospores have been observed in connection 

 with it on the gooseberry; but such is stated to have occurred on 

 the Continent, and hence it is classed under Pucciniopsis. It is 

 difficult to conceive that an iEeidium spore can survive from 

 season to season, and the inference from this would appear to be 

 that the relations of this parasite are as yet problematic. Again, 

 the wheat-mildew is by no means limited to districts where the 

 barberry is found, so that it is highly probable that that plant is 

 not the only one on which the secidiospores are produced. Another 

 instance might be adduced in Coleosjporium Senecionis, which is 

 frequently found in situations miles distant from fir-trees. Such 

 cases would appear to indicate that much still remains to be dis- 

 covered in this branch of investigation. Some errors of observation 

 have, besides, possibly taken place, owing to imperfection in the 

 modes of investigation, and to the obscurity of the relationships of 

 these low organisms. Notwithstanding all this, I think there can 

 be no doubt of the reality of the heteroecism itself in many of the 

 Uredines, and thus a new field lies before us for study and obser- 

 vation. As to the very restricted nature of the heteroecism, as 

 between two, or at most a very limited number of host plants, 

 there seem to be many facts — some of which have already been 

 referred to — that point to relations existing over a wider area. 



At the heart of the mystery of this heteroecism lies, no doubt, a 

 good and sufficient reason. The problem is hard; but you may, 

 perhaps, pardon me if I hazard a guess. Cluster-cups are never 

 very injurious to the plants on which they grow, while rusts and 



