PHYTOGEOGRAPHICAL REMARKS. 



493 



number of species. Probably continued study of the flora will increase this 

 number, but here also the difficulty of collecting complete material, is a 

 factor which must be taken into account. 



Of the smaller orders are conspicuous by the large number of endemic 

 species : Marantaceae with 5 out of 20 ; Aristolochiaceae with 3 out of 5 ; 

 Anonaceae with 5 out of ,20; Lauraceae with 5 out of 22; Rosaceae with 

 S out of 19; Lecithydaceae with 7 out of 16; Myrsinaceae with 4 out of 8; 

 Gentianaceae with 7 out of 21 ; Asclepediaceae with 5 out of 16; Lentibu- 

 lariaceae with 5 out of 12. 



Very small is the number of Podostemaceae compared with that known 

 for British Guyana. Of the 6 known species 2 are endemic (and new). This 

 small number is probably caused by the difficulties attending the collection 

 of these inhabitants of the water. Finally comparing the percentage consti- 

 tution of the flora according to the orders with that which is given by 

 Schomburgk, we find the following result : 



The agreement is fairly satisfactory. The greatest deviation in shown 

 by the Gramina which in Surinam occupy 5 % of the number pf species 

 and in British Guyana only 3 %. But we must take into account that in 

 1848 when Schomburgk's list was published, grasses were much less known 

 than nowadays, so that a new treatment of the representatives of this order 

 in British Guyana would probably considerably increase the number of 

 species. On the whole the age of Schomburgk's list is a factor which must 

 be take into account and which may cause the agreement in the percentage 

 constitution to be more accidental than real. 



Another conclusion which might be drawn from this comparison is that 

 it is not the savannahs of British Guyana which make the flora so much 

 richer in species than that of Surinam. For the orders that play an im- 



