EFFECT OF SUBSTRATUM ON DEVELOPMENT OF PARASITE. -17 



host, but soon die off, if it be not a suitable one. De Bary 

 also observed a germ-tube of Peronospora pygmam, which 

 IVeciuents Aiiemonr, making its way into Hammcuhis Ficaria, 

 l)ut soon to die. Germinating spores of Cystopus canduhts will 

 enter the storaata on leaves of any of their host-plants, especially 

 Capsella, but will only develop further if they are successful 

 in penetrating into the cotyledons. 



Variation in the substratum produces very great difference in 

 the formation of the reproductive organs. Thus many Ustila- 

 gineae produce conidia by continuous sprouting only when 

 cultivated in nutritive solutions, while their resting-spores are 

 developed only from a mycelium which inhabits the reproductive 

 organs of their host; this is the case with Ustilago caricis, 

 U. anthcarum, and U. tritici. In others the spores are found in 

 all parts of the tlower, and even in the inHorescence, as in 

 Ustilayo criwnta and If. tragopogonis, while in Ustilago maydis 

 spores are also produced in leaves and stems. 



The various parts of the same plant behave very differently in 

 this respect. The Ustilagineae just considered reproduce them- 

 selves only on certain organs of their host, although the 

 mycelium is also present in other organs. Other fungi behaving 

 similarly are UjncJdoe tyjjhina which produces its perithecia only 

 on the surface of the sheath of one of the leaves just below the 

 inflorescence ; Aecidimn datirmm develops its aecidia only on 

 the needles of the witches' broom; Aecidium cnphorbiac has its 

 aecidia only on the leaves of its host ; Ecoascus p)rnni has asci 

 only on the fruit ; Calyptospora produces teleutospores in the 

 epidermal cells of the stem, never of the leaves ; and so on in 

 many other cases. 



The formation of oogonia of Cystopus exhibits a striking vari- 

 ation according to the host-plant. Cystopus candidus on Capsdla 

 produces conidia alone, never oogonia ; yet the latter are plenti- 

 fully developed in flowers of Brassica, being confined, liowever, to 

 the flowers, while conidia are produced in all parts. Cystopus 

 hliti forms conidia only in the leaves, and oogonia only in the 

 stems of Amaranthus hlitum} 



The mycelium of many other fungi can only grcjw in certain 

 organs, while germ-tubes from the spores are only able to pene- 

 trate into certain parts of the host. Thus, E.coascus cdni incanae 



M3e Bary, Morphology and Bio/or/y of the Fun;//, English Edition, j). lid]. 



