92 Mr. R. B. Hinds on Geographic Botany. 



in the six great divisions or provinces is 1 in 325, and the amount 

 fluctuates in different countries on both sides of this estimate, ac- 

 cording to physical circumstances. Within parallel tropical regions 

 the amount is smallest, and increases as we advance tov^^ards the 

 north, through the subtropic and temperate regions, till the max- 

 imum is attained in the arctic. This progressive increase in the 

 northern hemisphere is no doubt due to the configuration of the 

 land, the large continental masses here closely approximating, 

 and forming almost a continuous surface about the arctic circle, 

 assisted also by the great similarity of climate. In the southern 

 hemisphere there is precisely an opposite distribution of land 

 and water, the continents gradually growing narrower towards 

 the south, and yielding to the ocean of waters, which at the ant- 

 arctic circle is scarcely broken by land. Of 233 species col- 

 lected in Kotzebue^s Sound one half are found in Europe, whilst 

 a similar proportion crosses Behring^s Straits, and are repeated 

 in Siberia. Though the affinity between similar regions in the 

 two hemispheres, or between distant mountain chains, is strongly 

 characterized, the relations of identity are extremely slight ; in 

 the latter particularly so, where it is rare to meet with species 

 identically the same as those of the plains. 



It is by these three methods of relation that the flora of one 

 region or country is to be compared with another, and an ana- 

 lysis established, conclusive and satisfactory, whence its import- 

 ance as an isolated flora, or compared with others, is ascertained. 

 Alpine vegetation judged by these characters loses some of its 

 importance, its relation being chiefly that of affinity, the species 

 belonging to genera whose maximum exists near the level of the 

 sea; hence peculiar groups, as genera and families, are very 

 rarely limited to them. The features of the vegetation of the 

 lowlands are repeated in accordance with controlling circum- 

 stances, marked and peculiar characters being seldom met with. 

 To illustrate more fully these different relations, we will sketch 

 an outline of the flora of the Sandwich islands, which, from their 

 solitary situation in a wide ocean, are well adapted for this pur- 

 pose. 



These islands, eleven in number, including two which are 

 scarcely more than rocks, stretch obHquely across a point inter- 

 sected by 21° N.L. and 157° W.L. They are distant 2900 miles 

 from America, 3500 from Asia, and nearly the same from the most 

 projecting part of New Holland. Numerous islands intervene be- 

 tween the two latter of these continents, but those towards the 

 Sandwich islands are mostly small and unimportant. The climate 

 is extremely equable and not disagreeably warm; in 1838 the mean 

 temperature was 77°, and the range of the thermometer from 85° 

 to 66°, being nineteen degrees. IMuch rain falls in some of the 



