68 Botanical Society of Edinburgh. 



other source of confusion is the practice of collecting specimens 

 without numbering them and the tree, trusting subsequently for iden- ' 

 tification to the memory alone, whereby a most unpleasant feeling of 

 uncertainty is produced. The changes in the form of the leaves, 

 and in the relative proportion of some of the parts of fructification 

 at diflPerent periods of growth, are often so surprising, that without 

 a mark of recognition, I should frequently have doubted whether my 

 specimens had been all collected from the same tree. Again, it is a 

 common practice to select for preser^'ation the largest and most 

 vigorous-looking specimens, in consequence of which an erroneous 

 idea of the average character is very apt to be produced. If an un- 

 usually luxuriant specimen be chosen, it should have a correspond- 

 ing label. Were those whose residence is fixed for a great part of 

 the year to give their attention regularly to this interesting tribe, 

 and above all, to set aside a portion of ground in their gardens for 

 the cultivation of the most intricate species, much of the uncertainty 

 which at present deters botanists from the study of the Salices 

 would probably disappear ; but the hasty collection of fragments in 

 flower, and above all, the un-identified addition of leaves, sen'C 

 only to perpetuate mistakes. If I might take the liberty of recom- 

 mending to others a practice which I have myself profitably followed, 

 I should advise that the specimens of ever}- Salix in a herbarium 

 (excepting, of course, species about which there can be no mistake) 

 should be such as to present one or more regular series illustrative 

 of the progressive development of the catkins, each set being taken 

 from the same tree at intervals during the fiowering season, — and 

 tliat at least two specimens of the leaves, gathered at different pe- 

 riods, should be preserved, so as to show the form of the stipules, 

 and the progressive alteration in the foliage; — also, that thin sections 

 of a catkin of each species, perpendicular to the axis, sliould be 

 gummed down, by wliich means the form of the ovarium and any 

 other particular respecting it — the length and pubescence of its stalk, 

 the nectary, the character of the axis, and the number of ovaria in a 

 given length of the spiral, could easily be seen without mutilating 

 the other specimens. The exact date also of each specimen should 

 be registered, whereby many ambiguities would be removed. The 

 willows, though numerous here (Audley End, Essex,), are not cul- 

 tivated to such a profit as they might be. Salix Russelliana, though 

 plentiful, is confounded M'ith S.fragilis ; nor is the bark held in any 

 esteem. The cserulean variety of S. alba has, however, been sold to 

 advantage to the makers of bonnet-shapes, as it is reputed not to 

 stain. To show the quick return which the arborescent willows 

 would afford, I may mention that a tree of S.alba var. ccerulea, planted 

 in 1815, at the end of nineteen years measured in circumference, at 

 one foot from the ground, seven feet eight inches ; and in October 

 1841, at twenty-six years of age, its circumference was ten feet three- 

 quarters of an inch, and its height seventy-seven feet." 



The author tlien proceeds to give a particular description of each 

 species in the above groups — remarking of S. decipiens, that though 

 it appca's to be c)f little use for economical purposes, " it forms a 



