Page:TheTreesOfGreatBritainAndIreland vol01.djvu/185

 in Cornwall, South Devon, and Gloucestershire, which is very near to, if not absolutely identical with, the Fontainebleau tree, as some of the specimens have leaves which resemble rather those of the varieties rotundifolia and semilobata. The South Devon tree produces fertile seed, which has been planted, and the offspring differs in no respect from the wild trees. In English trees the flowers are reported to have a disagreeable odour, and the fruit ripens in the end of October or November. When fully grown, but still hard, it is olivaceous brown in colour, with numerous scattered small brown or grey dots; but when quite mature it becomes reddish. At Minehead in Somerset, the Nightingale Valley and Leigh Woods near Bristol, and at Castle Dinas Bran, Denbigh, the variety decipiens occurs. Mr. E.S. Marshall observed a remarkably fine specimen with good fruit on the Conan river in East Ross-shire; but as no other specimen was seen this tree is probably not wild in that locality. The tree in Earl Bathurst's woods near Cirencester has given rise to some difference of opinion. It was identified at Kew as Pyrus intermedia; but in the specimens which I have seen' the leaves have the triangular lobing and tomentum of Pyrus latifolia, and I have no doubt that it is this species. Its foliage is very variable, some leaves being broad, with rounded bases like the type, whilst others have narrowed bases, and approximate in outline to the decipiens variety.

Pyrus latifolia is seldom planted except in botanical gardens, as at Kew, Edinburgh, and Glasnevin. There are several fine trees at Edinburgh, one of which was figured in the Gardeners' Chronicle for 1882, when it was 45 feet high by 5 feet 3 inches in girth. Professor Balfour had the tree measured again in January 1904, when it was 45 feet high by 6 feet 6 inches in girth. A year or two before it was considerably pruned on the top branches, and this probably accounts for it not being higher in 1904 than it was in 1882. Professor Balfour kindly sent me specimens of the Edinburgh trees, which, though they differ slightly, are all referable to Pyrus latifolia. He informed me that while the birds eat the fruit off one tree as soon as it is ripe, in another the fruit remains on the tree untouched. The variability of the fruit in this species is remarkable, and points undoubtedly to hybrid origin.

A tree exists at Oakleigh House, near Keynsham in Somerset, which was planted many years ago.