162 wight's botanical letters. 



this place ; while for want of my specimens I am occasionally 

 at a loss to determine, whether plants differing from the 

 descriptions, are species or varieties. This difficulty I have 

 often experienced among the PolygalcB^ partly it is true from 

 its being said in the generic character that the lateral lobes 

 of the corolla are always abortive, which is so far from being 

 the case, that I have now, I believe, as many as four species 

 recently collected, with the lateral lobes exceeding the alse. 

 In some they are so conspicuous when growing that I at first 

 sight took them for Crotalarias, the said abortive petals re- 

 sembling pretty large vexillums. Perhaps the generic char- 

 acter ought to have the word often inserted before '' abortive, 

 which might suggest a convenient division of the genus into 

 two groups, those with and those without lateral lobes or 

 petals to the corolla.* 



I have lately found Polanisia felina, and seen abundance 

 of Caparis divaricata^ but not one plant in flower. Of the 

 Malvacem I have as yet gathered only a few, but have found 

 wild, for the first time, the Hibiscus eriocarpus in dense 

 jungles near Nagary, a fine country for botanizing ; but un- 

 fortunately I was a few weeks too early to reap the full 

 advantages of the opportunity I had while passing through it. 

 I had not been there before, but would like to go again and 

 for a longer time. I have now got a large supply of Bytfne- 

 ria herbacea, and also of Lagunea lobata, the last not in a 

 good state. For the former I was a little too early, there 

 being very little fruit ; in respect to it, I made the following 

 note with reference to the stamens : — " Filaments 10, five ot 

 them sterile, somewhat ligulate, obtuse, sometimes anther- 

 bearing ; 5 fertile, each divided at the apex, and bearing two 

 one-celled anthers, or rather perhaps a single double anther, 



* At the moment of writing the above, Dr Wight appears to have misun- 

 derstood the generic character ; comparing Xanthophyllum with Polyg'^^'^' 

 it will be seen that the two symmetrical petals of the former are wantmg 

 in the latter ; that is, there are only three petals more or less combined m 

 Polygala {not Jive) ; moreover it is the second and third petals, not t e 

 fourth and fifth, which are abortive. — G. A. W. A. 



