ON THE SPECIFIC RANK OF POTAMOGETON ZIZII. 117 



considerable extent. Setting aside such doubtful forms as P. 

 coriaceus and P. varians, I found a distinct local type in each 

 series ; and although many of these local races varied to such an 

 extent that they often closely approached 1\ heterophyllus^ yet these 

 varieties retained their local facies. Now 1 know that this variation 

 in the fenland races could not in every instance be the result of 

 differing local conditions, because many of them grew under exactly 

 similar conditions ; and further, because one of them, which 1 had 

 cultivated in a small tub of water, retained its character for sonu 

 years. I arrested the growth of this plant some times, and en- 

 couraged it at others, and so contrived to get as many states as 

 possibJe ; but in the end I found these states exactly corresponded 

 witn tnose of the same ages in the wild plant. 



I then examined the fenland specimens of P. grawineus L. from 

 the same localities, and found that they varied to some extent, and 

 probably were not all true gramineus. In my note on this latter 

 species, I touched on the fact that this species seems to divide itself 

 naturally into two forms — the graminijolius and heterophyllus of 

 Fries. Not to complicate this paper too much, I will confine 

 myself to the one point of the difference between the forms of Zizii 

 that grew witn graminijolius and those that grew with heterophyllus. 

 In the first place I found that the Zim forms associated with 

 graminijotius were all of one type ; and altliough some of them 

 approached that species very closely, yet no corresponding approach 

 to Zizu was made by graminifolius itself — evidently the two forms 

 did not cross reciprocally, in this locality the Zizii forms are quite 

 away from the more usual fenland and southern forms, and show a 

 decided approach to the northern type of the species. 



The Zizii forms growing along with v. heterophyllus of Fries are 

 much more varied, tiie coriaceous floating leaves are much more 

 abundant, and the heterophyllus forms themselves often show a 

 great resemblance to typical Zizii in some of their states. With 

 little doubt the^e plants are reciprocally fertile. Probably 1\ 

 varians is a second cross of heterovhullus with Zizii. a double-hybrid 



of like character to P. coriaceus. Here we are in the region of 

 mongrels, plants that keep no constant character, but vary in a most 

 perplexing manner; many seedlings differing considerably from 

 tneir parent forms. 



P. Zizii has one other character which strongly points to 

 hybridity ; it varies from a sterility so complete that even flower- 

 spikes are never produced, up, by degrees, to a fertility equalling, 

 m the individual fruiting spike, and exceeding in the number of 

 spikes, tnat of P. Luc ens. Iffat the sterility of some individuals is 

 not due to unfavourable local conditions is proved by fertile and 

 barren plants growing side by side, borne local races of Zizii are 

 much more fertile than others ; in some localities a barren plant is 

 rarely to be met with ; this is especially the case when a fertile 

 colony is growing away from P. gramineus L., and is, perhaps, an 

 indication that the fertility of hybrid Potamogetons is not increased 

 by recrossing with one of the parents. Certainly the supposed 

 double-hybrid P. coriaceus is very much less fertile than typical 



