162 TRANSACTIONS OF ROYAL SCOTTISH ARBORICULTURAL SOCIETY. 
must be much more profitable to the owner than one in which the 
majority of the trees are of a bushy habit, or are liable to begin 
growth under the influence of short spells of warmth, as is often 
the case with larch. Carefully arranged observaticns might, there- 
fore, do good service to practical forestry by clearing up this 
question. These are a few of the objects of research which, in our 
opinion, have a distinct bearing upon practical forestry, and which, 
at the same time, require no expensive process for their fulfilment. 
Many others of a similar nature could be named, but enough has 
been said to afford an idea of the general scope of research work 
suggested. 
PRINCIPAL FEATURES OF AN IDEAL EXPERIMENTAL 
Forest AREA. 
In considering the various features which a suitable area should 
possess, we shall assume that it is desired to attain the whole of 
the objects named above. The first thing we have to settle is 
Geographical Position. 
This must be partly decided by the extent to which the forest will 
be used in connection with existing educational institutions, actual 
and possible. Considered as an appendage to the forestry classes 
in Edinburgh, the nearer to the latter city the better. But we 
take it that it is more desirable to first of all establish a forest area 
which shall combine the several functions already enumerated, and 
this being so, the needs of purely academic institutions cannot 
alone be considered, but also those of the majority of practical 
foresters, estate owners, etc., who would regard the area as a con- 
venient object-lesson and source of information. 
As a necessary condition for its position, we should consider as 
indispensable a sufficiently inland locality, out of the reach or 
influence of strong sea-breezes, which have an injurious effect upon 
many trees. Latitude is of less importance than elevation; but, 
considered as a national experimental area, the more it approached 
the central districts of Scotland the better, both in respect to the 
convenience of all concerned, and the collection of data for reliable 
yield tables. It need scarcely be said that proximity to a railway 
station is of great importance, for if situated in a remote and almost 
inaccessible district, its value would be greatly decreased. 
