Parasitic Funrji of Illinois. 393 



natural classification, the life history of plants must be taken 

 into account, no matter how obscure and impractical, for or- 

 dinary work, the characteristics it affords may be. It is prob- 

 able that De Bary's arrangement is more logical; but for 

 convenience most modern botanists use Leveille's classification, 

 which is the one adopted in this paper. 



The true position and rank of the group in the vegetable 

 kingdom still seems to be a matter of doubt, no two authors 

 exactly agreeing in regard to it. The arrangement adopted by 

 Winter (Die Pilze, II.) perhaps on the whole best expresses 

 their true relations. It is essentially the same as that proposed 

 by Saccardo (Sylloge Fung., I.) and may be expressed as fol- 

 lows: — 



Class Ascomycefes. 



Order Pyrenomtjcetes. 



Suborder Perisporiacece. 



Family Erysiphece. 



It is usually easy to determine, even without the aid of a 

 magnifier, whether or not a given fungus belongs to this 

 family; and a moment's examination of the mature perithecium 

 under the microscope will sufiice to place it in its proper genus. 

 Specific determination is often a matter of much greater 

 difficulty. In other groups of parasitic fungi, as, for instance, 

 the Uredinece, the species are, for the most part, confined quite 

 closely to a single host, or at least to a few very closely related 

 host species. Very slight differences in the form or markings 

 of the spores, taken in connection with a difference of host 

 plant, are considered of specific importance. Some of the 

 Erysiphea' have long been recognized as having a much wider 

 range of habitat; but it was natural for the botanist, if he 

 found a plant of this family developing on a new host, to 

 consider it a new species, especially if he observed a few more 

 or a few less asci or spores, or found the appendages differing 

 in number or length from the descriptions of other accepted 

 species. Then, too, owing to the poor instruments at their 

 command, the descriptions of the earlier investigators are 

 lacking in those microscopical characteristics that are now so 

 carefully noted. While the above view of narrow specific 



