94 PROCEEDINGS OF THE AMERICAN ACADEMY 



longiore ; labii inferioris trisecti lobis 3 fere conformibus obovatis iis 

 labii superioris bifidi minoribus ; ovarii loculis 3-4-ovulatis. — Collin- 

 sia grandijlora Hook. Kew. Jour. Bot. 3, p. 298, nou Lindl. Willow 

 thickets of the Valley of the Kooskooskee, in the western part of 

 Idaho, Spalding, Geyer. The sessile gland representing the fifth 

 stamen is at the very base of the corolla ; in the preceding species it is 

 higher up on the tube and smaller. 



Pentstemon barbatus Nutt., var. trichander. Humiiior e 

 caudice lignescente ; antheris longe parceque lanoso-barbatis ! — S. 

 W. Colorado, T. S. Brandegee, in Hayden's Exploration, 1875. Mr. 

 Brandewee was struck with this as different from P. barbatus in its 

 grdwth and asjiect ; but I see, no character to distingu'sh it from the 

 var. Torreyi of that variable species, except the long hairs on the an- 

 thers, and sometimes a few on the filaments. Tliis has not been else- 

 where met with in the Elmigera section. But it occurs with such 

 variability and apparent inconstancy in P. glaber and some allied 

 species, that it may not be relied on here. 



Pentstemon Clevelandi. P. spectabili quoad folia et inflorescen- 

 tiam baud dissimilis ; foliis superioribus arete sessilibus nee connatis, 

 floralibus minimis ; thyrso racemiformi nudo floribundo ; pedicellis 

 breviter filiformibus ; calycis parvi lobis ovatis capsulam 3-4-plo brevi- 

 oribus ; corolla sanguinea tubuloso-infundibuliformi (fere pollicari), 

 fauce paullo ampliata, lobis brevibus rotundatis jiatentibus ; filamento 

 sterili apice dilatato hinc barbato. — Cafion Tantillas in Lower Cali- 

 fornia, received from D. Cleveland in flower, and later from Dr. 

 Palmer in fruit. 



MIMULUS Linn. Having had occasion to elaborate the species 

 belonging to the Californian flora, I have thought it best to give a 

 synoptical view of all the known North American Mimull. Some are 

 difficult to limit, and the extent of the genus was also uncertain. 

 There are three or four groups of species, which would necessarily 

 rank as genera distinct from true Miimdus, if they were not connected 

 by transitions. Perhaps the most marked of these is represented by 

 two dwarf Californian annuals with* long filiform tube to the corolla, 

 and a cartilaginous capsule, the valves of which bear the half-septa 

 and placentas. One of them was probably the type of the genus 

 Jiunanus Benth. in DC. But the tube shortens and broadens in a 

 long series of species, which at length pass into true Jilimulus, in 

 which the placenta sometimes partially and rarely completely divides. 

 Equally peculiar in habit are the shrubl)y species or forms on which 



