OF ARTS AND SCIENCES. 259 



in most, from very base to apex. In the group of which E. dejlexum 

 is the type, the scar is more or less supra-basal (in E. hispidum 

 almost central and oval), and the apex of the nutlet is more or less 

 free beyond it. E. glochidiatum^ A. DC, which Beutham refers to 

 Paracaryum, is surely a true Eckinospermum, allied to E. Virginicum. 

 E. canum, Benth. in Royle, 111., wliich C. B. Clarke takes to be " the 

 type of the genus Echinospermum as described in the Genera Plan- 

 tarum," but which he refers to a glochidiate section of Eritricltium^ 

 is surely of the former genus. As to the exact position and length of 

 the scar, Bentham's character of his Entrichiece, "nuculis . . . apicibus 

 circa stylum plus minus prominentibus erectis liberis" applies to some, 

 but not to all, of the plants he referred to Echinospermum and to 

 Eritrichium. The importance given to this character and to the 

 corresponding height of the gyuobase in the Cynoglossoid genera 

 has not led to good practical results. The glochidiate armature of 

 Echinospermum and Cynoglossum may still be regarded as essential to 

 the character. 



Sclerocaryum is with good reason referred back to Echinospermum 

 by Boissier. Its stout spines are not indistinctly glochidiate at the 

 blunt apex. E. Sinaicum (^Eritrichium Sinaicmn, DC.) is apparently 

 a congener. 



Echinoglochin, Gray, Proc. Am. Acad. xii. 163, and Syn. Fl, ii. 

 190, I would still retain as a section of Echinospermum, coming 

 nearest to Krynitzkia. Also the section Homalocarynm, A. DC, now 

 containing a good number of species, should be kept distinct from the 

 section Lappida. In the supra-basal insertion of the nutlets of some 

 (but not all) of the species, as well as in habit, this group comes near- 

 est to Cynoglossum and to Omphcdodes. In the Lappula section the 

 published name E. Fremonti, Torr. in Pacif. R. Rep. xii.- 46, has been 

 accidentally omitted from Syn. Fl. ii. 190. It is a synonym of E. 

 EeJoivskii, var. occidentale. 



Omphalodes, Tourn. By leaving 0. linifolia^ 0. littorah's, and 0. 

 amplexicaulis in this genus, Boissier and Bentham seem to abandon 

 most of the ground on which the former's genus Paracaryum rests, 

 except as to the longer-flowered species. For the species of Ompha- 

 lodes above mentioned have a high-pyramidal gynobase, on which the 

 nutlets are borne, for the most part in an ascending position ; and their 

 attachment extends to below the middle of the ventral face. (The 

 nutlets carry away with them a "lacinula" of the style, in the 

 manner of Cynoglossum.') Rotate or short hypocraterimorphous 

 corolla, somewhat supra-basal or ventral attachment of nutlets, these 



