W. Gihhs on the Platinum metals. 341 



idseqiiante paullo brevioribus.— Either glabrous or hirsute. Richardson's 

 plant from the arctic coast is a large-flowered form of Torrey's P. alpina. 

 Parry has an alpine form (No. 287), and a loose, evolute form with longer 

 and narrower leaves (No. 284) ; in these the tube of the corolla is usually 

 pilose inside near the middle; but it is not so in Torrey's original speci- 

 mens Qi M. alpina, nor in Hooker's M. Drummondii. In the latter, and 

 m Parry's specimens, as in all of the various other species I have exam- 

 ined, the f^tainens are inserted in the throat of the corolla. In the flowers 

 of Dr. James' specimens, they are inserted pretty low down on the tube, 

 so that the tips of the anthers barely reach to the level of the faucial piiccC 

 or ;ippeiuiMgL's. This is the case in all the various specimens I have exam- 

 'n*-'d (of Buike, Fremont, and Slansbury) from the western side of the 

 Rocky Mountains, of what seems to be a narrow-leaved and hirsute vari- 



a short style. 



§ 4. Filamenta antheris sublongiora et equilata : corollse limbo lobato: 

 aeheniaechinata! 



.10. M. RivuLARis, DC. M. elUptica, Ledeb. ex Rege! & Tiling, Fi. 

 Ajan, N.E. Siberia and Karatschatka. Corolla with the tube hairy 

 Within towards tiie base; plica; at the throat conspicuous. I have only 

 a specimen from Tiling's Ajan collection. In this the fruit is conspicu- 

 ously echinate with soft prickles, — a remarkable peculiarity, which is not 



■"** Dr. Hooker, in his Arctic Essay,' received long since the above 

 ^s written, adopting Sir William's suggestion, refers the high arctic 

 ^- Drummondii {Lithospermum Drummondii) to our M. Virgimca. 

 Although Lfchmann describes tlie corolla '''fauce notata protuberantiis 

 fnnque, I found no appendages in an original specimen in herb. Torrey, 

 just as Dr. Hooker notes. But I also found them obsolete in specimens 



^^ Sierdbre I rejected the character from the diagnosis of Section 3. 



Art. XXX.~Besearches on the Platinum MetaJs; by WoLCOrr 

 GiBBS, M.D. 



(Continued from vol. xxxi, p. 71, Jan., 1S6I.) 



§3. 

 ^^'^- mass of soluble clilorids obtained by the above method 

 ■ > al 1 the plati nutn metals, ^although only ' 

 ■i;idi(un are present; \x\ 

 ^ -'luble matter, consist! 

 d' and partly of undissolved oxyds. A certam portion 

 ;iKso remains with the mass even after careful washmg. 

 ■\ashings contain a very large quantity of iron, a little 

 -'an ia the form of bichlorid of ruthenium and potassium, 



