126 



that on the Swedish mainland, oiie of the caiises being the long 

 and mild autumns enjoyed by the island. 



It may also deserve to be pointed out that J. D. Hooker, in 

 his work „The Studenfs Flora of the British Islands" 

 (London, 1884, page 123), states that the flowering-period olP.fru- 

 ticosa is only „June — July . 



* 



The greatest interest with regard to P. friiticosa is attached to 

 its character as dioecious. I shall here give a short description of 

 the different kinds of flowers I have observed, and also state the 

 result of the statistical investigations I have undertaken for the 

 purpose of discovering the proportion existing between the different 

 kinds of flowers. 



In the cT-flowers, the stamens are well developed and the pis- 

 tils highly rudimentary, or entirely wanting (see fig. 4). In the 

 9-flo\vers, the pistils are well developed, although they are some- 

 times only few in number, and the stamens are more or less rudi- 

 mentary, but are never entirely absent (see fig. 5). I have distin- 

 guished the following four kinds of flowers, based upon the varying 

 degree of the reduction of the stamens and the pistils: 



1) The stamens well developed; the pistils few and sterile or 

 none; the gynophore with a buncli of hairs (= a less strongly con- 

 firmed cf-flower, given in the table as ef 2). 



2) The stamens well developed; the pistils and the bunch of 

 hairs of the gynophore altogether wanting {= a most confirmed 

 cf-flower, given in the table as cT 3). 



3) The pistils well developed, although sometimes few, the stamens 

 large but sterile (= a less strongly confirmed 9-flo'^^'er, given in 

 the table as 9 2; see fig. 5 a). 



4) The pistils well developed, although sometimes few; the stamens 

 small and sterile (= a more strongly confirmed 9-flower, given in 

 the table as 9 3; see fig. 5 b). 



On the examination of 5,110 bushes from 24 different colonies, 

 it proved, that 1,895 of them, or 37,08 X, were cf-individuals, and 

 that 3,215, or 62,92 %, were 9-individunls. Of these 5,110 bushes 

 I made a closer examination of 3,210 from 9 of the colonies, for 

 the purpose of discovering the proportion existing between the four 

 stages of development I have defined above. I found that 25, or 

 0,8 %', belonged to (f2; that 1,166, or 36,3 %', belonged to the ef 3 



