124 



The Scottish Naturalist. 



reduced to a minimum. The permanent numbers, of course, 

 must not — unlike the tree-marks — be used for more than one 

 bush. 



Having attached to each example its number, any particulars 

 which can be seen more readily in the fresh than in the dried 

 plant may be entered in the note-book. These should include 

 the colour of the leaves, twigs, stigmas, and anthers ; the nature 

 of the filaments, — whether free, or more or less combined — 

 pubescence, if any, on them ; shape of the nectary ; style of the 

 venation of the leaves, whether raised or impressed, &c. The 

 specimens should be very carefully dried, and subjected to as 

 much pressure as will keep the leaves from wrinkling, but not so 

 much as to crush the catkins altogether. The leaves should be 

 so arranged, that the underside of some of them, — both the old, 

 or lower, and the younger, or upper, — be shown. 



The periods of growth when specimens should be taken is im- 

 portant. Male catkins should not be too old, but some should be 

 in full flower, and some with the pollen shed, so as to show the 

 colour of the empty anthers. Female catkins should be neither 

 too young nor too old, and should illustrate the condition of the 

 fully developed stigmas. When it is possible, specimens in fruit 

 should also be obtained, but not too old, as burst capsules are 

 comparatively useless. 



The leaves should not be taken till they have arrived at 

 maturity, and not, as a general rule, before the middle of August. 

 Young leaf-specimens are sometimes useful as supplementary 

 examples. It is important, in taking leaf-specimens, to secure 

 side-branches as well as shoots, since the character of the leaves 

 on these is often different. 



In collecting Willows every bush should be examined, not only 

 when in flower, but when in leaf, and, if there is the least doubt 

 about the species, specimens should be taken. Where more than 

 one species grows together, a sharp look-out should be kept for 

 hybrids, but hybrid forms sometimes occur at some distance from 

 either of their parents. This may happen by the transmission of 

 the seeds by wind or water, or by the parent bush, cross-fertilised 

 by insect-agency, having died out. In looking for hybrids it must 

 be remembered that frequently they do not show exactly inter- 

 mediate characters, but often bear a close resemblance to one or 

 other of the parent species. Such forms can be detected only by 



