insertion; rays fully developed and about 1>4 inches long, and 

 filiform pedicels 4 lines long. Bishop, California, May 15, 1897, 

 in the Lower Temperate Life Zone. 



Cymopterus aboriginum var. subtematus n. var. 



Like the variety ovalis, but leaves rather ternate than pin- 

 nate, and therefore deltoid in outline, final segments oblong- 

 oblanceolate, about 1 line long; fruit nearly orbicular. I would 

 not hesitate to call this a good species if there were not inter- 

 grades. Collected at the same time and place as the variety. 



Cymopterus aboriginum var. oblongus n. var. 



Like the type, but final divisions of leaves linear, 2-3 lines 

 long; fruit oblong, about 5 lines long and 2^^ lines wide, 

 the body linear and about >^ line wide, and about half as wide 

 as the lateral wings, which are almost always more developed 

 than the dorsal, cross section of wings linear throughout. 

 This appears very different from the type, but both grew near 

 together in the same locality and have the same peculiar seed 

 face. For years these forms have lain in my herbarium under 

 the name of C. cinerarius, but it is evident that they are not 

 that species. 



An instance of the viciousness of the Brittonian code un- 

 der w^hich they have been compelled to train is found in the 

 treatment Coulter and Rose give of Cymopterus purpurasceus, 

 and all on the account of one obscure varietal name. I pub- 

 lished Cymopterus Utahensis and its varieties monocephalus 

 and Eastwoodae, and C. purpurasceus, which will stand as the 

 proper names of these species and varieties. Coulter and Rose, 

 in spite of the description in Bot. Ives and on the strength of 

 a single specimen which may or may not be the type of the 

 variety change the names of all four in transferring them to 

 rhellopteriis, though the Vienna rules require that they 

 should not be changed. 



Leaving Pteryxia and taking up the other species of 

 Cymopterus, two courses are open to the investigator; the 

 Grayian and the Roseau. The former recognizes the difficul- 

 ties of arrangement, but follows the genetic relationship (al- 

 ways a safe guide) and groups them all under Cymopterus. 

 The latter method ignores the genetic relationship and ar- 

 ranges the species under four genera, thereby increasing the 

 confusion three fold or more, for when under Cymopterus the 

 difficulties can readily be understood by following the genetic 

 relationsh'p as shown by structure and ecotogic vl conditions 



