8 ]\Iuhlenbergia, Volume 8 



Leaflets with lobes and teeth normal, appressed villous; 

 plumes about 50 mm. long E. cinerascens 



13. Foliage hoary, villous-tomentose; calyx lobes long-acumi- 

 nate, surpassed by the linear bractlets E. grisea 

 Foliage pilose-ciliate; calyx lobes ovate-lanceolate, not sur- 

 passed by the bractlets. E. aliena 



The key is, of course, but tentative, and intended especially 

 for the use of field workers in the various localities where the 

 little known forms occur. The illustrations, will, of course, be 

 used in interpreting the key, average leaflets (from near the mid- 

 dle of the leaves) of most of the forms being represented. The 

 expression "ciliate-cuspidate," used above, may be taken to mean 

 a terminal tuft of fine hairs very much simulating a cusp. 



Additional observations. — Leiberg ^jjj (Oregon)and Kronk- 

 hite j8 (Oregon), are mentioned by Greene (12) as perhaps ref- 

 erable to E. canescens. Their pubescence surely suggests that 

 of E. canescens., as does also that of Kennedy i6jo (Nevada) and 

 oi Kellogg and Harford 20(p (California); but all four have the 

 cleft leaflets of ciliata proper, and three (Kennedy's plant being 

 in flower only) have also the ciliata style. 



The 39 fruiting specimens of E. ciliata that may safely be 

 designated as jointed are from: California i; Washington 4; 

 Oregon 6; Idaho 6; Utah i (my own /<57o); Montana 4; Yellow- 

 stone Park 2; Wyoming 6; and Colorado 9. The other 6 speci- 

 mens in fruit are: Shear 4422 Beaver canyon, Idaho; Afearns 

 2go'/ Yellowstone, and Rydberg and Bessey 4414 Bridger moun- 

 tains, Montana, which have the .style evidently unjointed, but 

 the terminal non-accrescent portion deciduous. Griffiths and 

 Hunter 80 Camas Prairie, Oregon, E. W. Sheubcr Sunset, — ? 

 and Rydberg 660 Custer, South Dakota, are evidently unjointed 

 and persistent. Rydberg 660., however, should perhaps have 

 been better classified as a transition form, and placed with the 

 following. 



As to E. trijlora^ the 16 fruiting specimens came from the 

 following states: Illinois 5; Wisconsin 2; Minnesota 3; Iowa 2; 

 South Dakota 3; and a specimen collected by Havard from the 

 "Upper Missouri." The leaf characters of this Havard speci- 



