6 Aluhlenbergia, \'olume 8 



Beginning with E. ciliaia^ 64 specimens are found with 

 corollas that may be judged as to color. Of these, 6 may be 

 called }-ellow with conspicuous red-purple apices, another 8 yel- 

 low with pronounced red (haidiy purple) markings, and 50 must 

 be put down as pale yellow \Aith more or less reddish margining 

 and veining of th.e petals. Those of E. cainpaiudata, as de- 

 scribed b}- Greene, are conspicuoush- 'Vuimson-tinted," which 

 agrees with what I have called "red-purple" above. Two of the 

 3 specimeus of \-ar. orjiata hax'e pronounced a])ical and margi- 

 nal red-purple, as also i out of 6 specimens of ^. dissecta^ which 

 may be passed ui)on. As to the other 5 specimens of ^''. dissecta 

 and those of brevifolia^ aiistralis^ and flaviila, worthy of men- 

 tion, the petals are simply pale yellow, as in most of the ciliata 

 material at hand. 



Of the 17 specimens of E. trijiora with flowers present, 9 

 show prominent markings with red or red-purple, 4 are appar- 

 ently almost unmarked pale yello\x-, and in 4 the petals are so 

 withered and discolored as to discount an\' coi;ciu>ion in their 

 case. Of 6 flowering specimens of E. ciuerascens, 3 ma\- well 

 be called "yellowish-purp'e," while in the other 3 the slight red 

 pigment present might easih- be neglected. Without making 

 detail records as to the rest of the Tnflorae^ since so few speci- 

 mens are at hand, I may simply sa)- that in all specimens of E. 

 affinis, arisonica^ [Hartiua?i jod), grista and aliena, in which 

 the color of the corolla mav be judged, it is the same |)ale yel- 

 low with more or less inconspicuous reddening of the \-eins. 

 The specimens of E. canescens and j^. tridental a gi\e us no cer- 

 tain information as to the corolla coloration; but I fully expect 

 that it will prove to lie the same with them as with the last four. 



Key to the described forms. — One of the fir.st things at- 

 tempted upon getting access to the de^^criijiion nf (Treene'.>> seg- 

 regates, was to make a ke\' for use in determining specimens, 

 especially ni)- own; but the result was so unsatisfactory that the 

 attempt was soon abandoned. With the specimens before me, 

 liowever, and tabulations made, the following ke}' was con- 

 structed, items from Greene's descriptions being used as freely 

 as possible. 



