102 



size from 4.5-8.5 cm. broad (the one in Fig. 14 is 8 cm. broad) and 

 are either soUtary or in clusters of two or three. Sometimes 

 there is alternation from year to year in the inflorescence. Fig. 

 13 shows very marked clustering of fruit, while the following 

 year the flowers were solitary. 



The calyx-appendages are either short or long. In the plant 

 of Fig. 13 from Jones Island the appendages are foliaceous, 4-6 

 mm. wide, and, like the leaflets, serrate and glandular beneath. 

 The shape of the leaflets and their very coarse serration give 

 this specimen a peculiar character. An interesting variation is 

 shown in Fig. 14 where two of the sepals are lobed. Such forms 

 are not rare at Elgin, and occur chiefly in the earliest flowers. 



What attracts even the casual observer of the Elgin plants 

 is the variation in the shape of the fruit. Typically, I suppose, 

 the fruit is subglobose or slightly obcompressed. Such is the 

 prevailing form at the coast; but obovoid, ovoid, deltoid, oblong, 

 elongated and strongly flattened forms are very common. 

 Curiously enough at Elgin these various forms often occur pure 

 or almost so; and thus give character to large clumps (Fig. 3). 

 This is in no way due to environment, as several very marked 

 forms alternate in one habitat. Neither is it due, as I at first 

 thought, to the great number of carpels matured in one and 

 the small number matured in another. Whether they represent 

 new forms in process of development or old forms hopelessly 

 confused I cannot say. 



On the dike of the Serpentine River near Elgin several of 

 these forms grow together. First were a few low bushes with 

 hispid fruits (var. hispida Fernald). Next came a hedge of 

 globose forms; and then, without a gap or change in the char- 

 acter of the clumps, plants practically all the fruits of which 

 were obovoid, longer than wide and with an acute base (Fig. 3). 

 More globose forms were succeeded by strongly flattened forms, 

 the body of which was almost twice as broad as long (Figs. 7, 8, 

 9) ; while at the far end of the hedge were some not very healthy- 

 looking plants with elongated necked fruits not reddening well 

 on account of some fungus, though the nutlets were well formed 

 (Fig. 5). Out on the diked flats were globose fruits with or 



