466 THE SALIX LISTS IN THE 'LONDON CATALOGUE.' 1399 cinerea L. Smith's two varieties aquatica and oleifolia are no more than leaf-forms connected with the type by inter- mediates ; the names may be of use both for the garden and the herbarium, as an aid to classification of the forms of a very variable willow, but they do not represent distinct varieties ; and on this ground were omitted by Dr. White. It has been suggested that these two represent hybrids of aiirita; it is not the case, however, that all ^'aquatica is aurita and Caprea, nor that all oleifolia Sm. is aurita X cinerea, as has been suggested. After deducting what may rightly go to these hybrids, there still remains a residuum in each case of pure cinerea forms. 1401 Caprea x cinerea. I am still unable to give a satisfactory account of the distribution of this hybrid. It is apparently ex- ceedingly scarce, and hard to detect even in localities where both species occur together. A plant cultivated at Kew as iS^. moschata would seem to be it ; but the plant formerly cultivated there as S. sphacelata was in my opinion S. aurita x Caprea rather than this hybrid. A Derbyshire form of it is issued as No. 55, Set of British Willows. 1402 repens. The varieties of 8. repens admitted in former editions of the Catalogue can only be regarded as forais of little constancy (see introductory remarks to the third fascicle of British Willows) ; and do not cover half the forms which this species pro- duces in Britain. The only familiar form which is a good variety is S. rosmarinij'olia L., and that is perhaps rather a subspecies ; while the proof of its occurrence in Britain is not forthcoming. 1402 repens X Caprea (Caprea-repens Lasch.). Published by Dr. White for Britain (Journ, Linn. Sac. xxvii. 394) on our Armadale specimens, but I have never agreed to this naming; and now after years of cultivation (at Shirley), and after seeing Scandinavian specimens of Caprea x repens and cinerea x repens, I am sure that my view is correct, that the plant is S. cinerea x repens. This latter hybrid comes also from the Clova Valley, but not from Holme 1^'en, as Dr. White thought on specimens of the Rev. W. E. Linton's or mine in Mr. A. Bennett's herbarium. I have also a leaf- specimen, probably of this, received from Dr. Mason, origin unknown. We have, however, no S. Caprea x repens for Britain as yet. 1403 phyliclfolia X aurita (S. ludijicans F. B. White). The account of the Perthshire bushes on which S. ludijicans was founded leaves room for grave suspicion whether any of them were S. aurita X phylicifolia. I have come across several such bushes, with foliage much on the phylicifolia side, and aurita suggested by the pubescence and shape of the leaf, and it may be an abbreviated style. Culti- vation, however, in two such cases disposed of the pubescence, and has shown the plant to be S. phylicifolia. The most promising Perthshire plant died. The most satisfactory Forfarshire plant is issued as No. 59, Set of British Willows. There is in all probability some strain of aurita in this last ; but Dumfriesshire plants, of which Mr. James Fingland has kindly furnished me with cuttings and specimens, which are admirably intermediate, show how mucii wanting in evidence of aurita the Perthshire ludifcans is. All such
