especicilK' of the difficult faiiiil}' GrainiiKac, where the last mono- 

 graph is more or less closel)' followed. In many cases this was 

 a very commendable way, but in others not, as, for example, in 

 the treatment of Poa, where he follows Professor Beal. One 

 improvement he has made on the latter's work, viz., in retaining 

 Poa P^ucklcyiina and P. Fcndlcriana and their allies in Poa. He 

 places them under a subgenus Atropis, copj'ing Real's characters 

 of the genus Atropis [wWxch name however does not belong there 

 but to Piicciiicllid), but not noticing that scarcely one half of the 

 species referred there by Beal agree with the definition ; nor did 

 he know that Atropis Lctta-manni Beal {^Poa Lcttcrvianni Vasey) 

 and Poa Braiidcgci described in Real's work are the same species 

 and that the t}'pes of both were collected at the same station. 



As no work has been published before on the flora of the 

 region, Mr. Howell had to draw his information from a thousand 

 and one scattered publications. We know that many times the 

 same species has been described under different names by differ- 

 ent authors (one Aster from Idaho, A. Jcssicac, has received not 

 less than four names). A good deal of sifting had therefore to 

 be done and it is remarkable how well Mr. Howell has succeeded 

 without having access to the types. It would be surprising, 

 however, if he had not gone amiss sometimes. One such case 

 we have noticed : Sporoholiis gracilliiiuis and 5. filiforuiis were 

 both based on ]''ilfa dcpaupcrata v. filiforniis Thurber, and hence 

 the same. 



The numerous publications and segregations of recent date have 

 of course caused considerable trouble. Some of our contempo- 

 rar\' phytographers have a custom of describing species without 

 indicating the relationships. The author of a monograph or 

 manual, if he does not have the chance of seeing the types, must 

 be a very good guesser if he happens to place the species in the 

 right section of the genus. Mr. Howell guessed well as a rule, 

 but missed occasionally, as, for example, when he placed Gcn- 

 tiana anisosepalcx Greene, next to G. affi)ns. It should have been 

 placed with G. tciiclla and G. acuta. 



Another kind of difficulty arises when one of the modern 

 splitters breaks up a species, supposed to be transcontinental. 



