76 



apices more or less acutely pointed; anthers about or almost i 

 mm. long, alternating with the corolla lobes ; filaments apparently 

 connate basally, or at least broadly attached to the corolla. 



Found in the miocene shales of Florissant, Colorado, by one of 

 the University of Colorado expeditions. Named after Mrs. 

 Marion Durbin ElHs, who first recognized its generic position. 

 It probably belongs to one of the two species already described 

 from Florissant, based on leaves, but it is given a separate name, 

 owing to the impossibility of referring it definitely to either of 

 those species. It is a very pretty little flower, remarkable for 

 the regular form, with the anthers exactly in place, this being 

 due, no doubt, to the short and broadened filaments. The 

 corolla lobes are more pointed, with straighter sides, than in the 

 modern species known to me. 



I take this opportunity to note that Mrs. Cockerell has suc- 

 ceeded in uncovering the basal leaflets in the type specimen of 



Fig. I. Samhucus Ellisiae. FiG. 2. Phalaris (?) geometrorum. 



Samhiicus amahilis Ckll., and they prove to be trifoliolate, as is 

 not rarely the case in the modern species. The lateral divisions 

 are narrow, and considerably smaller than the apical one. 



Phalaris (?) geometrorum sp. nov. 



A lemma 19 mm. long, nearly 3}/^ broad near base; apex nar- 

 rowly acuminate, but not awned; margin not hispid; median 

 nervure colorless, hardly at all visible; outer nervures (one on 

 each side) thick and dark, reddish-black as preserved, doubtless 

 dark green in life, but pallid apically. 



