22 
P. fennica. At the time when I arrived at the conclusion that 
Mr. Craig- Christie’s Arran Pyrus was P. scandica, and distinct from 
the English plant so called, I drifted into another error — that all the 
pyrus of the Aria group found in Arran ought to be referred to P. 
scandica , supposing that the Arrau Pyrus with the leaves pinnate at 
the base was a plant described by Pries as P. scandica , var. pinnatifida. 
At that time I had seen very few, and these few imperfect, specimens 
from Arran, and possessed but a single Scandinavian specimen from 
the late Dr. Blytt. Having, however, obtained good specimens of 
P. fennica collected by Dr. Ahlberg, of Upsal, some of which agreed well 
with Mr. Duthie’s specimens of Pyrus from Glen Catacol, I began to 
think that the supposed P. scandica pinnatifida from Arran must be 
P. fennica , and in order to resolve my doubts I made an excursion to 
Arran in the first week of June, 1872, and found both partially 
pinnate and the non-pinnate forms of Pyrus. As soon as I saw the 
plants growing I was convinced that the non-pinnate and abundant 
plant was true scandica , and the scarce form with leaves pinnate at 
the base was true fennica. P. fennica differs from scandica in having 
all the vigorous leaves with from 1 to 4 pairs of the lowest lobes 
separated quite down to the midrib, and the succeeding pair, or even 
two pairs, nearly so. Prom each pair of separated pinnae above the 
first there is a decurrent stripe on each side of the midrib, broadest at 
the point where it leaves the upper pinna, and decreasing in width till 
it vanishes at the origin of the pinnae beneath it. The first pair of 
pinnae is almost always as long as and no broader than the succeeding 
pairs on the adjacent unseparated lobes if there be but one pair of 
pinnae. The part of the leaf which is not cut into separate pinnae 
is ovate or rhompic-deltoid or triangular towards the apex, with lobes 
decreasing in size towards the apex. The separated pinnae are oblong- 
oval or elliptical, usually remotely and coarsely serrate on both margins, 
often appearing acute from the prominence of the terminal tooth into 
which its partial midrib runs, but when this tooth is smaller than 
ordinary the outline appears rounded. The lowest of the lobes which 
are not separated so far as they are free resemble the pinnae. The 
succeeding ones, which are much shallower, are usually more acute 
towards the apex ; besides, the main veins which run into the pinnae, 
or the extremity of the lobes, vary from 6 to 10 on each side, but 
besides these there are often intermediate lateral veins which run into 
the sinus between two lobes, so that the number of lateral veins is often 
much greater than that of the lobes. The upper surface of the leaf 
even in its young state has not a distinct flocculent covering, but has 
arachnoid hairs, especially along the veins, as in P. Aucuparia. The 
