258 PROCEEDINGS OF THE AMERICAN ACADEMY 



Lehm., the nutlets of wbkh separate from the very low-pyramidal 

 gynobase by a manifestly lateral scar. 



A few notes upon the principal genera concerned may serve as an 

 introduction to the following systematic exposition of certain mainly 

 North American Borraginacece. 



Cynoglossum, Tourn. Some recent authors have ignored a character 

 (peculiar to this genus and to a part of Omphalodes) which was well 

 delineated by Schkuhr almost a hundred years ago, and noted by 

 Alphonse De Candolle in the Prodromus, namely, the carrying away 

 by the apex of each nutlet of an exterior portion or " lacinula " of the 

 indurated and persistent style, by which the four nutlets, after splitting 

 off from the gynobase from below upward, are for a while suspended. 

 Although in some species the nutlets are nearly horizontal on a de- 

 pressed gynobase, in others they form a low or even a considerably 

 elevated pyramid, with corresponding elevation of the gynobase ; and 

 the areola of attachment, or scar, varies from " si/prci medium " to 

 very manifestly infra medium of the ventral face. The fruit of C 

 ccelestinum, Lindl., would refer this species rather to Omphalodes. 



Paracaryum. of Boissier does seem to be a too heterogeneous and 

 ill-limited an assemblage. As founded in 1849, and as maintained in 

 the Flora Orientalis (1879), it includes Lindelojia of Lehm., 1850, 

 although the character calls for included stamens, which is not quite 

 true of the latter. It also calls for a funnelform corolla, with tube 

 more or less elongated ; and upon this character the genus apparently 

 may stand, whether including or excluding Lindelojia, Lehm. Ap- 

 parently this should be included ; for the anthers emerge more or less 

 from the throat of the corolla in P. heliocarpum, P. angustifolium or 

 azureum., &c. But whether Boissier's name, given to a wider group, 

 or Lehmann's stricter but rather later name, should be adopted, is a 

 nice question, which T am not called upon to determine. The sjiecies 

 with short corolla will, upon this view, fall back into Omphalodes, 

 Rindera.. and Echinospermum. 



Echinospermum, Lehm. It seems hardly correct to attribute this 

 genus to Swartz, simply because (as Lehinann states) he suggested the 

 name, while the published species with which his name is indirectly 

 connected belong to Cynoglossum. It would appear that a better view 

 is taken of this genus by Bentham in the Genera Plantarum than 

 bv C. B. Clarke in the Flora of British India. The latter would in 

 effect restrict the genus to the Lappnla group, in which alone, or 

 mainly, are the nutlets attached up to their apex. In some of these 

 (as E. Redowskii) the nutlets are " attached above their base " ; but 



