12 CGN.TRIBUTIOXS TO WESTERN BOTANY NO. 15 



i^goes Mm one better.' It sb happens that most of the species treated by 

 Kydberg are familiar to me in the field. To found species on the extent 

 of pubescence, and' the length of petals, shows a great lack of knowledge 

 of the plants as they grow where the hairs vary from long and slim ,to 

 short and stubby, which leads Robinson to use the term granular, an 

 inappropriate one, for the short and stubby hairs. But Rydberg carries it 

 tarther. An examination of the species of Lepidiura reveals gre' t 

 va/iance in the shape of pods of the same species. Then it is unfortu- 

 15 ate that some other character was not used in describing pods as winged 



a ad in grouping them according to the extent of wings, which often sepa- 

 rates related things. The result is that the species of Lepidium are still 

 m a hopeless state of confusion, tl does not do for us to dismiss the 

 question as Robinson does when he does not accept L. alyssoides as a 

 \ar. of montanum by saying that though the species may intergrade at 

 ceaairr localities they are still distinct. Extreme forms can be separatea 

 Ix't in the field it cannot be done. Alyssoides, in the form called Jonesii 

 hv Rydberg is shrubby as in L. Fremontii in the St. George region, but 

 thfs character varies to simple biennial. The persistence of woody forms 

 counts for nothing in the hot regions, where even tomatoes live from year 

 fo year and become bushes. All these things serve to emphasize the more 

 the patent fact that closet botanists are not equipped to monograph 

 genera. Tn my judgment these inatters never will be settled till they are 

 taken up by plant breeders. I consider Robinson's work in the Syn. Fl. 

 ?s the best yet done by closet botanists, but there is still room fox 

 improveraent ; • 



It is impossible, to follow Thellung in his treatment of Lepidium. 

 He does not know the species he treats. He makes a new species ofijiy 

 L. pinnatifidum from San Diego and calls it L. Robbinsonii. 



L. Virgihicum- varies from forms with large to minute .petals or absent:' 

 The leaves seem mostly to be spinulose-toothed, all other characters fail' 

 The pedicels are mostly long, 2-3 times the pods. I cannot make ari}'thing. 

 out of L. ramosum and ramossisimum Nelson. L. Vaseyanun ' and Cran- 

 d"alli' seem to be scopulorum. L. crenatum fs montanum. L. divergens is 

 a form of apetalum. 



Rydberg p. 19 Cont. N. Y. Bot. Card 287 does not spell San Ber- 

 nardmo correctly nor Astragalus Bemardinus. 



In commenting on my subdivisfons of Astragalii Rydberg brin^^s out 

 what Ke considers inconsistencies, and they undoubtedly exist, but no one 

 can treaf any genus without being more or less inconsistent at times, for 

 plants persist in growing in their o^to way. It. is true that A suc- 

 cumbens is not very near to the Malaci, and was long ago put by' Gray 

 and Watson in a separate group which they called the Succumbentes, but 

 this arrangement does not solve any relationships. I . for a long time keot 

 rp. even- almost to the last, thfs grouping but in the end became convinced 

 mat genatic considerations made my last arrangement better and I still 



