POLEMONIACE^. 51 







cylindrical, 3-4 lines long, the segments scarcely spreading. Monte Diablo Range and 

 northward. 



ASCLEPIADACEJE. 



Herbs with milky juice, no stipules, and regular flowers, with the parts in fives, 

 except that there are two carpels with distinct ovaries and a common stigma to which 

 the stamens are attached; the latter (in our genera) with hood-like appendages: leaves 

 entire, generally opposite, sometimes whorled: flowers usually in simple umbels: fruit 

 a pair of follicles. Seeds almost always with a coma of silky down. Key to genera 

 and species, p. 142. 



ASCLEPIAS. 



In Bay -Reg. Bot., A. Californica, Greene, is Gomphocarpus tomentosus, of this 

 book, A. ecornuta is Gomphocarpus cordifolius, and instead of Schiznotus, 

 Greene writes Solanoa. The latter plant grows near the Geysers of Sonoma Co. 

 where it was first collected by G. B. Towle and on the mountains north of Clear Lake. 

 Its ved flowers and often prostrate habit make it a noticeable plant in the order. It 

 may well be called Towle's Milkweed. 



GENTIANACE^E. 



Glabrous herbs, with colorless, bitter juice, entire, opposite and sessile leaves: no 

 stipules, perfect and regular flowers: stamens as many as the lobes of the corolla and 

 alternate with them, inserted on the tube, the anthers free from the stigma: ovary 

 1 -celled: style one or none: the stigmas commonly two. Calyx persistent. Key to 

 genera and species, p. 143. 



POLEMONIACE^E. 



Chiefly herbs with simple or divided leaves, and no stipules: all the parts of the regu- 

 lar flower five, except the pistil, which has a 3-celled ovary and a 3-lobed style. Calyx 

 imbricated in the bud, persistent: corolla convolute in the bud: stamens on the corolla 

 alternate with its lobes, distinct: anthers introrse. In Gilia the cells of the ovary arid 

 the stigmas are occasionally reduced to two. Key to genera and species, p. 145. 



It is very difficult to define the genera of this order. If we arrange all our species 

 in groups according to their affinities these groups interlace more or less deeply. In 

 other words, one or more species are common to two or more groups. Therefore, when 

 we separate these groups under generic names there are species that might as well be 

 put in one as the other of adjacent genera. Before many of these troublesome inter- 

 mediate species (connecting links) were known, botanists easily made out a dozen or 



