1898] OPEN LETTERS 437 
Other similar instances may be found in the List of Pteridophyta ana 
Spermatophyta in Northeastern North America (1894), particularly under 
vex. 
A reasonable rule, which would avoid these changes, would be: Only 
identical combinations shall be considered homonyms. 
CORRECT NAME ACCORDING TO 
NAME PROPOSED BY SHELDON. PRESENT WRITER 
Astragalus viridis tmpensus. A. v. elatus (Wats.). 
A. shetrocarpus curvicarpus. A. s. falciformis A, Gray. 
A. preussti laxispicatus. A. p. laxifiorus A, Gray. 
A. p. arctus. A. p. latus Jones. 
A. leucopsis curtus. A. 1. brachypus Greene. 
A. franciscanus longulus. A. f. virgatus (A. Gray). 
A. megacarpus prodigus. A. m. parryi A, Gray. 
A. sparsifiorus majusculus. A.s. major A, Gray. 
A. glabriusculus spatiosus. A. g. major A, Gra 
A. atratus arctus. A. a. stenophyllus Jones. 
It will be noted that Mr. Sheldon himself gives the same varietal name 
to two species. This may bean oversight, or it may be that he considers a 
varietal name invalid only when used (if not under the same species) in a 
specific sense elsewhere in the genus. At the same time, he changes a 
varietal name when the alleged homonym is a pure synonym, so long as it is 
a binomial. 
A curious case is that of A. crotalari@ var. virgatus Gray. 
both names are preoccupied in a specific sense, so Sheldon calls the species 
A. franciscanus. According to my views, Gray's varietal term virgatus may 
be retained as varieta/, though it cannot be applied to the species, because of 
the earlier 4. virgatus Pall. Thus we get A. /franciscanus var. Miia g 
an instance of a varietal name older than that of the species.—T. D. 
COCKERELL, Mesilla Park, N. M. 
It seems that 
