1891.] PUBLIC DOCUMENT — No. 33. 211 



squashes from J^.Ir. S- F. Libby of Milford, Mass., who re- 

 ported it as killing the leaves completely, and thus stopping 

 the growth of the plant and its fruit. It is vmquestionable that 

 farmers and gardeners have here a new and serious hin- 

 drance to the successful cultivation of squashes and cucum- 

 bers ; and there is no reason why it should not extend to 

 melons also. The fungus has never been figured, so far 

 as I am aware, and some facts in regard to its structure are 

 yet lacking. It is, therefore, thought worth while to sup- 

 ply some of these deficiencies here, with figures, for com- 

 parison, of the mildew of the wild star cucumber, which 

 may also, very possibly, be found on cultivated Ciiciirbita- 

 cecB in the future. 



The vegetative threads of P. Cuboisis and of P. a its/ ra- 

 il's do not difter essentially from those of other downy 

 mildews, and ramify among the cells of the host plant, 

 sending into these cells at intervals absorbing organs or 

 kausto7'i'a, by means of which they abstract nourishment 

 from them, and weaken or ultimately kill them. Among 

 the downy mildews we meet two t3'pes of these haustoria. 

 In the grape-vine mildew and others, they are small, knob- 

 like outgrowths from the fungus threads, merely large 

 enough to reach the interiors of the cells of the host ; while 

 in others, like the mildews of the turnip and of spinach, 

 mentioned later, they are large and branching, and often 

 nearly fill the cells within which they are developed. In 

 both of the species under discussion the haustoria are of 

 the first type, as shown in Figs. 14 and 16. The haustoria 

 of P. australis have been said to gi'ow often larcje, some- 

 times filling the cell, but I have observed none such. 

 From the vegetative threads grow out through the pores 

 in the epidermis of the leaf, chiefly on its under surface, the 

 threads which bear the suinmer spores or conidia. Those 

 of P. Cubcnsts are quite scattered over the yellow and 

 dead-looking spots caused by the development of the 

 fungus in the leaf, and do not form a close felt visible to 

 the naked eye, as do so many mildews, because it is rare 

 that more than two of the conidial threads issue from 

 a given leaf pore (Fig. 12). P. australis, on the 

 contrary, forms dense white tufts of small extent on the 



