578 Proceedings of the Royal Society 
Kamtschatka may be compared with those of North Norway, or 
Hakodadi in Japan with Vienna in Austria. 
On the 30° range line we find the north coasts of British Columbia 
comparable with those of North Norway, Paris with Fort Vancouver 
in Oregon, and the south of Spain with the north of Mexico; and 
lastly, on the 20° range line, Macao Island near Hong Kong in 
China with the same climate as Florida in America. 
A point worthy of notice, but for the establishment of which few 
data are as yet to be had, is that of the apparent diminution of 
range on elevated parts of mountains or mountain chains. The 
best authenticated example of this is that presented by the Alps 
of Switzerland. Observations for ten and twenty years at the St 
Gothard and St Bernhard hospices, at elevations of above 6000 
feet, give a range of only 27° for each, whilst the plain of Italy 
to the south has a range of upwards of 40°. This diminution for 
elevation is also observed in stations in the Pyrenees, the Tran- 
sylvanian Alps, and the Himalayas. 
By the aid of these charts of annual range we can at once predicate 
of any point of the earth’s surface whether it has a uniform and even 
climate, or an extreme one, or what precise place it holds between 
these limits. These charts may thus be of considerable utility in 
themselves, but it is when taken as companions to the annual iso- 
thermal charts, that they have their highest value. From the 
isothermal chart we may find that any required place has a certain 
mean yearly temperature ; but we have no means of ascertaining 
from it how far the temperatures of its coldest month may descend, 
or how high its warmest month may ascend the thermometer scale. 
Again, from the range chart taken by itself, we can only tell 
that the climate of the place in question is limited to a certain 
number of degrees, without being able to say what position these 
degrees occupy in respect to heat and cold. But let the charts be 
used in company, and then we may have all that is required. 
First, from the one chart find the mean annual temperature of the 
place, and from the other its annual range ; then, since the mean 
annual temperature of places in the temperate and arctic regions of 
the globe corresponds very closely with the midway point between their 
January and July temperatures — if the one-half of the amount of 
the range be placed above and the other half below the mean annual 
