48 RELATION OF PARASITE TO SUBSTRATUM. 



has a mycelium only in catkin- scales ; Exoascus pnmi, however, 

 hibernates in the twigs, and forms reproductive organs only in 

 the walls of the ovary ; Aeddium strobHinum grows only on the 

 cone-scales of spruce; Claviceps frequents only the young ovaries 

 of cereals and grasses; and so on, other fungi inhabit- only leaf, 

 stem, root, or flower. 



In this connection points of considerable interest are presented 

 by the behaviour of many Uredineae hitherto little investi- 

 gated. As was pointed out by De Bary, the germ-tubes 

 prodixced from both uredospores and aecidiospores (in Fiiccinia 

 dianthi those from sporidia also), penetrate into the stomata 

 of any phanerogamous plant. If, however, that should not be 

 a host-plant of the fungus in question, then the germ-tubes 

 die away in the stomatal air-cavity. If the host suits the 

 fungus only in a limited degree, then no hypertrophy will 

 result, and the latter will attain only to the formation of spermo- 

 gonia. Let the host, however, be the one best suited to the 

 fungus, then hypertrophy will result and aecidia be developed. 

 Very conclusive evidence of this interesting condition has been 

 furnished by numerous experiments which I have carried out 

 with spores of Gymnosporangium} If one infects Crataeg^is 

 Oxyacaifdha with G. davariaeforme, very marked stem-hypertrophy 

 results, even by the time the spermogonia have made their 

 appearance; there is also considerable swelling of leaves and 

 slight enlargement of cotyledons, while aecidia are produced 

 in numbers everywhere. When the same fungus is used to- 

 infect Pyi-us Awnoparia, no yellow spots or malformation 

 of any kind results, and spermogonia, hardly visible with a 

 lens, are formed only here and there. A similar infection on 

 Pyrus latifolia (P. Aria X torminalis) results in a crop of 

 badly developed aecidia. If quince be infected, then without 

 any hypertrophy whatever, little red spots bearing numerous 

 spermogonia are formed on the leaves, but the development of 

 the fungus ceases there ; on the death of the quince leaves 

 the chlorophyll is retained in the immediate neighbourhood of 

 the spermogonial spots, so that they remain for some time as- 

 green islands on the yellow leaf. E. Hartig's infections with 

 Melampsora tremulae also led to varied results ; on Finiis there 

 ensued a distinct disease of the cortex (Caeoma pinitorqmim)^ 

 ' V. Tubeuf, CentrcdUatt f. Bacterioloijic lu Parasitenkunde, 1891. 



