ITINERARY. Ixi 



And tliese remarks hold good of tiie s<lopes below aLso : so that Bnrk- 

 hill's regret, already quoted, has real cause. 



In this connection it will be of interest to mention some of the 

 special features brought out by Burkliill in his Introdviction (loc. cxt.) 

 as to the nature of the Flora of Roraima above 5000 feet. Referring 

 to the high proportion of endemic species (50*6 per cent.) among the 

 flowering plants, he writes : — 



'• The proportion of endemic Spermophyta may seem large, but does 

 not exceed that on record for some of the mountains of Mexico. Of 

 far greater interest is the number of endemic genera. The law that 

 mountains by their isolation and extension, as well as by their latitude, 

 produce endemic genera, is illustrated by their number in the ranges of 

 Cis-Equatorial (South America : thus there are eleven among the Sper- 

 mopliyta of Roraima, and onl}"- two in the Coast Andes, which are 

 comparatively small and not isolated, but I am aware of no fewer than 

 thirty-six in the extensive Andes of Colombia, including Mith them 

 the Cordillera of jNIerida. 



" The endemic genera are enumerated on p. 7. They belong to as 

 many orders. One of them — Heliamjjhora — has no kindred in South 

 America, but belong.*?, like Cyrilla, to a Xorth American group; all 

 the rest have more or less close allies in genera of the South American 

 continent. Ledothamnus, liowever, desex-ves further remark, because it 

 is one of the veiy few Ericace?3 with ericoid leaves which exist in the 

 New World." 



He points out also the special relationship of the Roraima flora with 

 that of the Andes, and not Avith that of the southern mountains of 

 Brazil, with which it has been compared, and in which any similarities 

 noted are much more due to similarities of climate than to any 

 evidently intimate relationship between them in times past. Thus he 

 names six genera as common to Roraima and the mountains of South 

 Brazil, but not reaching the Chilian Andes, being in fact not montane 

 in a restricted sense : also seven genera of the Andes which extend 

 from the north to Chili, also to Roraima and the mountains of South 

 Brazil ; while there are fourteen genera of the northern Andes, not 

 oveipnssing Boliva southwards, Init reaching Roraima, and not South 

 Brazil — besides thirteen other Andean genera extending to the Coast 

 Andes of V'^enezuela, reaching to the neighbourhood of Caracas, but 

 not yet found on the Roraima Group. 



A luminous passage may also be quoted from the Rejiort by 

 Stephani : — *'Tho collection of Liverworts made by Messrs. McConnell 

 and Quelch is a sn)all one, but is of particular interest from a geogra- 

 phical point of view Many plants were found which hitherto had 

 been only observed in the Andes of South America; their unexpected 

 appearance on the top of Mt. Roraima is quite startling : the curious 

 Frullania mirahilis, Jack et Steph., is of particular interest, as well as 

 the v^eiy rare and beautiful Fleurozia paradoxa, Jack, both of which up 

 to this time liad not been elsewhere observed." The collection of 

 Liverworts contjiined 40 species. 



Clearly therefore one may siun up the matter by saying that recent 

 extensions of oiu- knowledge of I he Flora of the Roraima disti'ict — in 

 which iMr. McConntll's .special e.xpcditiun boie an important part — but 



