WITH NOTES ON OTHER SPECIES. 87 



some chemical change has taken place. Shice the pits are marked bj colorless areas 

 in the otherwise deeply colored wall, the original membrane must have become so 

 changed, at those points at least, that it remains uncolored by the reagent. It seems 

 probable that the whole outer membrane assumes this character, and that the color is 

 produced entii-ely in the secondary deposit. 



The male organs, or aiitheridia, are, when present, almost always borne on slen- 

 der lateral antheridlal hrayiclies. In some species they are very rarely or never devel- 

 oped (Figs. 43, 104). Where they are present, they are not necessarily found in 

 connection with all the oogonia (Figs. 40, 75, 76), though they may be invariably so 

 (Figs. 52, 71, 105). The number of antheridia attached to a single oogonium shows 

 very little tendency to definiteness, except in A. racemosa (Fig. 94). The most that 

 can be said is that in a given species there is a general tendency to an abundant pro- 

 duction of antheridia, or the reverse. 



The antheridial branches arise from the main filaments oi* from oogonial branches, 

 sometimes exclusively from one or the other, in other species from both. They may 

 be very short and simple (Fig. 98), or long and much branched (Figs. 51, 71). 

 From their tips antheridia are cut off by transverse walls, and rarely are intercalary 

 also. The antheridia are cylindrical or clavate in form, somewhat thicker than 

 the branches, and with slightly denser protoplasm. Their form is very constant for a 

 given species and, in connection with their position and origin, affords important 

 characters for specific diagnoses. Their walls show in a less marked degree the 

 reaction of the oogonial walls. 



The same primary filament may give rise to both oogonial and antheridial 

 branches, or it may produce only those of one sort. Since it is practically impossible 

 to determine in ordinary cultures whether two filaments belong to the same plant, 

 and since definite cultures from a single zoospore have not been made, we cannot say 

 whether species whose hyphae are unisexual are truly dioecious or not. It is safe, 

 however, to apply to them the term used by DeBary, diclinous (Figs. 51, 88). The 

 same author calls species with bisexual filaments androgynous. Among species of 

 the latter kind, which constitute the large majority of the Saproleg niece, the anthe- 

 ridia may attach themselves to oogonia from the same or from other filaments. Most 

 of them reach oogonia from the same hypha, probabl}'^ because these are nearer. 



In one species not yet met with in America, 8. hypogyna, a single branch is 

 bisexual, the antheridium being formed from a cell cut off by a second wall immedi- 

 ately below the oogonium. This is the only known case in this family of antheridia 

 without antheridial branches. 



A. p. S. — VOL. XVII. L. 



