354 



NATURE 



[August q, koo 



glomerata, Alopecurus pratensis,%LQ.. (see Fig. i). (3) Tritici,on 

 Triticum vulgare (Wheat). (4) Airae, on Aira caespitosa. 

 (S) AgrosHs, on Agrostis stolonifera, &c. (6) Poae, on Poa 

 compressa and P. caesia. 



2. Puce. Phlei-pratensis, Er. at Hen., Aecidium unknown. 

 On Phleum pratense and Festuca elatior. 



3. Puce, glumarum (Schm.), Er et Hen. (Yellow Rust), 

 Aecidium unknown. Sp. form (l) Tritici, on Wheat (see 

 Fig. 2). (2) Seeahs, on Rye. (3) Hordei, on Barley. (4) 

 Elymi, on Elymus arenarius. (5) Agropyri, on Triticum 

 tepens. 



4. Puce, dispera, Er. (Brown Rust of Rye), with Aecidium 

 Aiuhusae. On Rye. 



5. Puce, triiicina, Er. (Brown Rust of Wheat), Aecidium 

 unknown. On Wheats — Triticum vulgare, eompacfum, spelta, 

 and dicoccum. 



6. Puce. broinina,'Ex., Aecidium Mvikviovin. On many species 

 of B ramus. 



wm 



:ii« 



% 



11 w' 



III 

 mi 





iilif 



Fig 2.— Yellow rust of wheat (/^7Cfc/«/W ^/«;«rt;ricw, spec, form Triticiy^ 

 a, leaf (nat. size), i^, outer glume (x 2), and c, a grain (x 2), bearing 

 the Uredo stage ; r/, uredospore ( X 375) ; e, sheath bearing Puccinia 

 stage ( X 2); /, a healthy grain, and g; a rusted grain (both X 2) ; 

 7i, teleutospore (x 500). (J. Eriksson 1 



7. Puce, agropyrina, 'Ex., Aecidium unknown. On Triticum 

 repens. 



8. Pucc.hokina,YjX., Aecidiu7n\yci}^r^oviVi. OnHolctis lanatus, 

 and H. mollis. 



9. Puce. Triseti, Er., Aecidium unknown. On Trisetuni 

 Jlaveseens. 



10. Puce, sifiiplex (Kleb.), Er. et Hen., Aecidium unknown. 

 On Barley. 



11. Puce, coronifera, Kleb., with Aecidium Catharticae. Sp. 

 form (i) Avenae, on Oat. (2) Alopecuri, on Alopecurus pra- 

 tensis, &c. (3) Festucae, on Festuca elatior. (4) Lolii, on 

 Lolium perenne. (5) Glyceriae, on Glyceria aquatica. (6) 

 Hold, on Holcus lanatus and Z^. tnollis. 



12. /'«ft:.6-(?r^Ma/a(Corda),Kleb. (Crown Rust), with .<4^«(//«w 

 Frangulae. Sp. form (i) Calamagrostis, on Calamagrostis 

 ariindinacea, &c. (2) Phalaridis, on Phalaris arundinacea. 

 (3) Agrostis, on Agrostis stolonifera, &c. (4) Agropyri, on 

 Triticum repens. (5) Hold, on Holcus lanatus and .^. tnollis. 



NO. 1606, VOL. 62] 



Comparing these species with those known in 1890, Puccinia 

 graminis is now divided into Eriksson's i and 2 ; Puccinia ruhigo- 

 vera into species 3 to 9 : the variety simplex is now the species 

 10; and Puccinia corotiala is divided into li and 12. 



The species are distinguished by characters of uredospore and 

 teleutospore, and by the host-plants of the Aecidium, where 

 known. A species may be subdivided into " specialised forms," 

 which agree in all external characters with the species, but form 

 (i) is tied to one or more species of host, and its uredospores 

 will not infect the hosts of forms (2), (3), &c. Thus Puccinia 

 graminis has a specialised form confined to wheat, one to oat, 

 and a third which attacks both rye and barley. The occurrence 

 of the Aecidium on the same host, e.g. that of Puccinia graminis 

 on barberry, might seem to afford a stepping-stone between 

 specialised forms; yet Eriksson says definitely that, for example, 

 in the form Avenae " the form of Aecidium on barberry which 

 gives black rust on oat can infect oat only." It may be men- 

 tioned that specialised forms are not peculiar to the rusts of 

 Gramineae. Klebahn in a recent paper (H. Klebahn, " Ueber 

 den gegenwartigen Stand der Biologic der Rostpilze," Botan. 

 Zeitung, 1898, pp. 145-58) gives a list of heteroecious fungi, 

 which show subdivision into "biological species" or "species 

 sorores," forms identical with or only slightly different from 

 those indicated by Eriksson'stermof specialised forms. Klebahn 

 has also an interesting discussion on the nature of these forms of 

 parasitic fungi, and joins issue with many of Eriksson's results. 



Carleton {loc. cit.) carried out a long series of experiments for 

 the United States Department of Agriculture. The procedure 

 was almost identical with that of Eriksson, and the results gen- 

 erally support the Swedish observations. Carleton distinguishes 

 the following species and forms^on cereals : — 



Orange leaf rust of wheat, Puccinia tubigo-vera tritici; 

 identical, or almost so, with Eriksson's Puccinia triticina. 



Orange leaf rust of rye, P. rubigo-vera secalis ; either 

 Eriksson's P. glumarum secalis or P. dispersa. 



Crown rust of oats, P.coronata,Coxda.; Eriksson's P. coronifera, 

 Kleb. 



Black stem rust of wheat and barley ; P. graminis tritici, 

 Er. et Hen. 

 ,, ,, ,, rye, P. graminis secalis, Er. et Hen. 

 ,, ,, ,, oats, P. graminis avenae, Er. et Hen. 

 The American results differ from the Swedish in two import- 

 ant points : — 



(i) Whereas Puccinia glutnarum tritici is the most destruc- 

 tive rust throughout Europe, it does not seem to occur in 

 America, where the common rust on wheat is Puccinia rubigo- 

 vera triticit 



(2) The specialised form Puccinia graminis tritici in America 

 appears to infect, not only wheat, but also barley ; this is not the 

 case in Sweden. 



These differences indicate that the rust of wheat or other 

 cereal in one country may not be the same as that in another ; 

 and that specialised forms are variable. Eriksson distinguishes 

 degrees of specialisation and classifies the forms thus: (la) 

 Forms restricted to one species of host, or to several species of a 

 genus, e.g. species 4, 8, and i (5) in the above list. {\b) Forms 

 occurring on hosts belonging to different genera, e.g. species 

 1(1) and I (2) in list. 



(2) Forms, which, under certain conditions of environment, 

 are less fixed in regard to their hosts ; thus, until recently, 

 Eriksson included species 4, 5, 6, 7 as forms of one species, 

 Puccinia dispersa, and it is still doubtful whether they are 

 distinct enough in their external characters to be regarded as 

 species 



Further experiments are necessary to clear up many points 

 regarding the distribution of species of rust. Yet the practical 

 bearing of the results is evident. For example, wheat cannot 

 now be regarded as subject to infection from any cereal or 

 grass showing rust, but only from those which are host- plants 

 of the species or specialised forms peculiar to wheat ; thus a 

 comparison i"rom the above lists shows that no rust from oats 

 has been found to infect wheat. The theoretical bearing of 

 biological species or forms in relation to the classification and 

 phylogeny of fungi is also of the deepest interest. 



The propagation of the rusts of cereals has, in the hands of 

 recent investigators, assumed new aspects. The life-cycle of 

 Puccinia gratninis as defined by De Bary, and already given 

 above, is the one generally described in text-books. Many 

 observers, however, have doubted whether this, the perfect 

 life-history, is always followed entirely. Strong objections of 



