[ 60 ] 



Herb. Amboynensis apparently on the supposition of Us being 

 indigenous in that island" 



It is needless to comment on this passage in detail, 

 but it is essential to observe that the plant with which 

 " Agave americana " is being compared is the Agave vivipera 

 (sic) of Linnaeus, and that it is for introducing a figure of 

 Wight's OWD Agave vivipara that Rumpf is (unjustly as it 

 happens) indited. The passage seems to have been often 

 read as if the plant of which Burmann added ( from Rumphius' 

 " Auctuarium) " a plate to the original account of " Semper- 

 vivum" was ' Agave americana.' 1 Actually it is neither A. 

 1 americana' nor 'A. vivipara' (of Wight)}mtA~ Cantala Hoxb 



Wight's illustration is not very good, but it cannot fail to 

 be recognized by any one who is conversant with the Indian 

 Agaves, and it should be equally impossible to confound either 

 the species itself or Wight's artist's drawing with the plate in 

 Rumphius. There is an unpublished drawing in the Herba- 

 rium H. B. C. marked as A. vivipara., L. in Wallich's hand- 

 writing under which is written " ? A. Cantula, Roxb." The 

 species is unmistakably that of Wight's figure, the more so 

 that both represent plants which have blossomed, when the 

 appearance of the " rosette " or bunch of leaves (which in 

 "vivipura" is highly characteristic) differs conspicuously from 

 the previous condition. 



Wallich's query was due probably to the original " Cantala " 

 having been lost from the garden ; and it must, we think, 

 be answered in the negative. 



In the first place, there is a plant which is fairly abundant 

 and self-propagated in some parts of India answering to the 

 figure and description in the Herb. Amb. which Roxburgh after 

 mature consideration cited for his Agave Qantula. 



In the second place, the Calcutta garden name for Wight's 

 species is " A. Contula var. vivipara" which suggests that the 

 Hoxburghian plant was readily distinguishable. 



The plant which we identify with A. Cantala is proved by 

 epecimens to be plentiful about Bombay, and through the 

 kindness of Mr. I. H. Burkill, F.L.S,, we have seen undoubted 

 examples of this well-marked species from the Northern Oircars, 

 the scene of Roxburgh's own chief explorations. 



Boughiy speaking, A. Cantala extends from Central India 

 (both E. andW. Coasts) to the N. W. Gangetio Plain and 



