ON SOME BRITISH VIOLA FORMS. 227 



form, but with the colour and texture of canina. Totally sterile in 

 the wild state, it has remained so for two seasons in the Reigate 

 garden. In the spring of the current year I noticed a fine capsule 

 growing among the numerous sterile flowers ; when going to 

 examine this capsule to see whether it contained seed, in the 

 middle of June, it was at once seen, from the foliage, that the 

 capsule was not borne by the hybrid, but by another plant which 

 had grown up in the middle of it. In order to prove this beyond 

 doubt, the two plants were dug up and carefully separated. The 

 new plant was a young example of V. Fdviniayia ; in fact, a seedling, 

 with the usual sterile rosette and two branches, each of which bore 

 a single capsule. I mention this particularly, because it shows that 

 the steriUty of the hybrid had no connection with its situation ; 

 and also that the visits to it of fertilising insects were not wanting, 

 as it is not conceivable that an insect should visit these two (the 

 only) flowers borne by the Iliviniana plant, without visiting some of 

 the numerous flowers of the hybrid which apparently proceeded 

 from the same plant. — Surrey, several localities ! ; Sussex, E. ! ; 

 Kent W., G. Nicholson \ 



V. CANINA Reich. — Apparently this plant should bear the name of 

 V. Jiavicomis Smith, who separated it from the sylvatica aggregate ; 

 but the examination of this point is incomplete. I am unable to 

 separate -vars. Jiavicornis Sm. and ericetorum Reich., except in their 

 extremes ; they seem to be states due to situation, and to pass into 

 each other. The Jlavico mis is a plant of heaths and commons where 

 the peat is close and compact, with a fine short growth of herbage. 

 The var. flavicomis, No. 22 of the Fasciculus, from the habitat 

 •'campus militibus exercentibus," is evidently from a locality such 

 as that just described; it is called " f . simjjlex" ; but the numerous 

 specimens are all of them pieces of plants ; if a plant be dug up, and 

 the hard peat carefully picked out (not an easy task), several of 

 these apparently simple pieces will be found to be united by their 

 rootstock into a single plant, branching below the soil much in the 

 usual way. The var. lucorum Reich. I know only by a specimen 

 from Cambs., leg. A. Fryer, determined by Herr Murbeck. The 

 appearance of this plant conveys to me the impression that it is a 

 state due to situation, rather than a well-marked variety. The var. 

 crcissifoUa Gronvall is a fine, thick, and somewhat fleshy-leaved 

 large-flowered plant ; yet it, too, passes gradually into the usual 

 fenland plant, which is itself a more robust form than that usually 

 found in the southern counties. — Cambridgeshire, A. Fryer \ 



V. LACTEA X CANINA ( F. Idctea var. intermedia Wats.). — From 

 studying this plant in its native habitat, and comparing the plants 

 with V. lactea and V. canina, which grow with it, I early came to the 

 conclusion that it was a hybrid between these two ; and this con- 

 clusion has since been amply confirmed by many observations. 

 This plant has somewhat the aspect of F. canina var. lucorum; 

 indeed, that name has on one occasion been applied to it. Observing 

 it to be seemingly sterile, attention was directed to that point in 

 1886 and 1887, the wild plants being examined at various different 

 seasons ; but not a single capsule was found either as the result of 



q2 



