356 NOTES ON THE ROOGEE OF KUMAON. 
twice as long, and very narrow. So it appears, at least from the spe- 
cimens in the Hookerian Herbarium, received from Bunge and from 
Karelin, as well as from Ledebour’s Plate 372 of his illustrations of 
the ‘Flora Altaica,’ although I cannot find that these fertile flowers 
have ever been described. Meyer, when he wrote the description for 
the ‘Flora Altaica,’ had only seen the minute sterile imperfect flowers, 
.. which come nearer to those of the Roogee in the shape of the petals 
. and sepals, and it is probably after having written the description that 
he inserted the reference to the above-quoted plate. Ledebour, in his 
* Flora Rossica,’ copies this reference without further allusion to iur: 
flowers; and neither in the * Flora Altaica’ nor in the ‘ Flora Rossica 
is there any reference to Plate 380 of the same illustrations, which gives 
a beautiful representation of the plant in fruit. 
— The stamens of M. polyandra have been already alluded to; the 
filaments are much thicker, and the anthers rather larger than ‘in A. 
laciniata. In the M. bifida (of which however I have only opened two 
flowers) they are less numerous, and more like those of the Russian : 
species. The ovary is rather more sessile in the Himalayan than m 
. the Russian species. The structure of the pod is, as has been already 
observed, the same in all three, but the shape differs: in M. polyandra 
each half is more regularly orbicular than in M. laciniata, and hori- 
 zontally spreading, the upper and lower edge being nearly similar; 10 
the M. bifida the pod is scarcely at all emarginate below, the lobes are 
much elongated, and although spreading at first, are curved upwards 
_as the pod ripens, leaving a very narrow sinus between them, and re- 
presenting a flat silicule split into two to about two-thirds of its length. 
- The structure of the seed is the same in all, except that the radicle is 
much shorter in the two Himalayan species, and especially in the Jf. 
. Any further detailed description of M. polyandra is rendered unne- 
cessary by those already given by Colonel Madden and Mr. Moore, in 
Dr. Balfour’s above-quoted notice of the plant in the Proceedings of 
the Botanical Society of Edinburgh. I therefore merely subjoin its 
technical specific diagnosis, together with the character of the hitherto 
unpublished M. bifida. : 
zacarpea polyandra, Strach. et Winterb.; caule elato, foliis pinnati- 
sectis, segmentis lanceolatis dentatis subincisis, panicula inermi, 
is petaloideis petala superantibus, staminibus multiplicatis (10- 
