258 PROCEEDINGS OF THE AMERICAN ACADEMY 



In describing this remarkable species, neither Scheele nor Buckley 

 mentions the annual root, nor the prevailingly alternate leaves, nor 

 the numerous ovules ; indeed, poor Scheele gives the character " loculis 

 uniovulatis." Dr. Engelman has proposed to transfer this species to 

 Gilia ; but the number of ovules in these genera proves to be a 

 wholly secondary character. 



2. COLLOMIA, Nutt. mutatis mutandis. 



Corolla tubuloso-infundibuliformis vel hypocraterimorpha, fauce 

 ssepius sensim plus minus ampliata. Stamina fauci vel infra faucem 

 incequaliter inserta : filamenta gracilia, saspe exserta. Ovula in loculis 

 solitaria, pauca, vel plurima. Semina humefacta e teguraento simpliei 

 tubulos mucilaginosos plerumque spirilliferos creberrime protrudentia. 

 — Herbae annua?, raro biennes, foliis alternis imisve oppositis sa3pius 

 incisis nunc pinnatisectis. 



Of the two characters which in the Prodromus distinguish Colhmia 

 from Gilia, namely, the unequally inserted stamens and the solitary 

 ovules, Bentham gave evident preference to the latter, as appears 

 from his removal of C. lieteroplnjlla to Navarretia ; yet uniovulate 

 species are left in Gilia. As it is now abundantly evident that none 

 of our Polemoniaceous genera can be made to rest upon the number 

 of ovules, I rely so completely upon the remaining character that I 

 propose to remove from Gilia to Colhmia two multiovulate species, 

 in which I detect a striking inequality in the insertion of the stamens, 

 and even to add an unpublished species having a much-dilated throat 

 to the corolla. 



Collomia nudicaulis, Hook. & Arn., has very many ovules, and 

 belongs to the Leptosiphon section of Gilia, although peculiar in its 

 sessile anthers and entire leaves. 



The " mucilage" so copiously developed on the surface of the seed 

 when immersed in water, and which gave name to the genus, consists 

 of innumerable and most delicate diaphanous tubes, which lengthen 

 wonderfully when wetted. The spiral thread which they contain (on 

 which account they were confounded with " spiral vessels," and which 

 uncoils as the tube softens or dissolves into jelly) is wanting in one 

 species, namely, C. gracilis. In this and in the several following spe- 

 cies, the mucilage cells are beneath a more or less evident pellicle or 

 epidermis, composed of fragile tabular cells, which are thrown off 

 when the former develop and protrude under moisture. But this pel- 

 licle is not obvious in the typical species. 



