T. Holm — Studies in the Cyperacece. 329 



with the terminal spike mostly gynaecandrous and the 

 TernaricB with the spikes fasciculate. And this progres- 

 sive development of the species coincides, to some extent, 

 with the geographical distribution: the simplest type 

 being represented in the boreal, the most evolute in the 

 southern regions. From the circumpolar C. suhspatJia- 

 cea the development gradually advances towards south, 

 culminating in the Ternarim: Ceylon, New Zealand and 

 Straits of Magellan. 



A corresponding development of the species may be 

 traced also to some of the sections themselves. For 

 instance in the Salince, when we compare the inconspicu- 

 ous, few-flowered (7. suhspathacea with the southern 

 variety of 0. salina, Kattegattensis, and with C. halo- 

 phila. Among the Cryptocarpm we pass from the boreal, 

 slender forms of C. cryptocarpa to the more southern, 

 very robust C. cryptocMaena. But regarding the Crini- 

 tce there seems to be no transitional types between C 

 maritima from the northern coasts and the southern, syl- 

 van types C. crinita and C. gynandra; nevertheless these 

 three species are undoubtedly closely related, and as 

 demonstrated above, they exhibit a remarkable variation 

 in the distribution of the sexes as well as in the composi- 

 tion of the inflorescence, imitating the most evolute types 

 of the genus. 



"W^iile thus the distigmatic species of the Aeoras- 

 tachycE have reached the most evolute stage of the genus, 

 exemplified by the Ternarice, the tristigmatic -do not 

 extend beyond the so-called ordinary type of Carices 

 genuince, except the peculiar C. Magellanica with the 

 habit of the Limosae, but with the spikes typically 

 gynaecandrous. 



Clinton, Md., August, 1921. 



