28 NORTH AMERICAN FAUNA. [No.16. 
possible to follow the uppermost trees wherever they may lead, but a 
map showing such a route would resemble a saw from which alternate 
teeth had been removed, the remaining teeth indicating the way the 
dwarf trees push up on the summits of ridges, the broad spaces 
between the teeth, the treeless gaps, usually the intervening valleys 
or basins. Trees always occur at some point in the bottoms of these 
valleys, and usually extend completely across them, but at an altitude 
a thousand feet or more lower than that reached on the ridges, and 
there is a material difference in the trees themselves. If of the same 
species, those in the valleys are much larger and taller; if of other 
species, as is frequently the case, they belong to the upper part of the 
belt below—the middle forest belt. On Shasta, the trees that push up 
highest on the ridges are always the dwarf white-bark pines, while 
as a rule those that bridge the intervening valleys below are full- 
grown alpine hemlocks or Shasta firs, the upper limit of which must 
not be mistaken for timberline. The difficulty lies in determining 
what ought to be considered true timberline, and the reason why 
in the absence of obvious barriers the white-bark pines do not fill more 
than a third or a fourth of the belt to which they properly belong. 
If a mountain could be found whose upper slopes form a true cone 
instead of a series of alternating ridges and valleys, so that suc- 
cessive transverse sections would be circular in outline, instead of 
irregularly scalloped, it is probable that timberline would form almost 
a true circle around the peak, rising a little on the southwest and 
dipping down a little on the northeast. But in the absence of such 
ideal conditions, actual visible timberline is usually confined to the bor- 
ders of the tongues of dwarf trees that occupy the summits of the radi- 
ating ridges (pl.1v). The explanation of the absence of trees from the 
intervening valleys is not always easily found; still, if the valleys are 
studied with reference to the details of their several slope exposures 
and other local conditions, the position of the hypothetical timberline, 
in most cases, will be obvious. Let us take, for instance, one of the 
numerous glacial basins on the south side of Shasta, bordered on each 
side by lofty ridges which are capped by tongues of white-bark pines. 
The bottom of the valley, whenever its axial slope is steep enough to 
be regularly swept by avalanches, can not, of course, contain trees. The 
broad basin slope of the ridge on the west faces east and is in its own 
shadow in the afternoon; as a consequence it is too cold for trees, but 
is well sprinkled with alpine plants. Its summit is covered with dwarf 
white-bark pines, which come up from the other side and end abruptly 
along its eastern crest. The cold eastern slope is, in its zone position, 
actually above timberline, although the tongue of dwarf trees along 
its crest may stretch up a thousand feet above the lowest alpine 
plants. 
On the opposite or eastern side of the basin the slope faces west or 
southwest, and receives the warm rays of the afternoon sun, The 
