BOTANY. 147 
would regard these as tardy stragglers in a southern migration. The phanerogamic 
flora of Central America (as calculated in 1888) comprised an intermixture of northern 
and southern types of varying proportions in different areas and an autochthonous 
element. The last constituted only eleven per cent. of the genera against seventy 
per cent. of the species. The percentages of genera restricted to America and of 
wider ranges were 53°7 and 46:3 respectively. For species the figures stand at 
89°9 per cent. restricted to America and 10°1 per cent. extending beyond America. 
The statistics of a specimen of the upper mountain-flora comprise 260 genera, of 
which eighty-two (or 31°6 per cent.) were restricted to America, thirteen (or 5 per 
cent.) endemic within our limits, and the rest, 165 (or 63:4 per cent.), had extra- 
American extensions. 
_ Nine of. the exclusively American genera had northern extensions ; twenty-five had 
southern extensions; twenty-two were common to the Andes only; and twenty-six 
were also represented in both North and South America. 
The total number of species enumerated is 604, whereof 504 (or 83-4 per cent.) were 
reckoned as endemic; 83 others (or 13:8 per cent.) not extending beyond America, 
with a residue of only 2°8 per cent. of wider range. The extensions to other parts of 
America are given as: N. America, 17; 8. America, 19; Andes only, 39; N. and 
S. America, 8. Of course, most of these figures are rough approximations, but they 
are sufficient to show that the representatives of northern and southern types are 
nearly balanced, and that the Andine element in the mountain-flora of Central America 
preponderates over the temperate northern element *. 
‘Taking the whole phanerogamic flora of Central America, as it was known in 1888, 
638 of the genera had north-western extensions, 562 north-eastern connections, whilst 
1285 had southern extensions. Species yielded similar proportions. But statistics 
alone explain very little. | | 
The present complexities of plant-distribution point to greater and oftener repeated 
alterations in the distribution of land and water than is generally admitted, and no 
one theory is sufficient, in my estimation, to account for the origin and progressive 
dispersal of organisms. Indeed, it is doubtful whether sufficient evidence still exists 
to carry us to a convincing conclusion. ‘There is always the great question whether 
organic as well as inorganic matter has not developed on the same or similar lines in 
different regions or centres. Comparing the following particulars of the distribution 
of families, it is difficult to realize that one part of the world has produced a land- 
vegetation wholly different in composition from:that of any other part. Very diverse 
interminglings exist, but they give little or no clue to the beginnings. | 
- Accepting, for purposes of comparison, the number of families of flowering plants 
at 290, as defined in the seventh edition of Engler’s ‘Syllabus,’ very nearly three- 
fourths are represented in both the Eastern and Western Hemispheres. In the more 
* Of, Biol. Centr-Am., Botany, iv. p. 299 (1887). 
v2 
