820 Yapp. — Spiraea IJlmaria , Z., and its Bearing on the 
generally flowers towards the end of July (or earlier on drier soils), by 
which time all the leaves on the shoot are unfolded. 
Thus the leaves of the erect flowering shoots vary markedly in 
hairiness, according to both their time of unfolding and their position on 
the stem. 1 
(c) The non-flowering shoots of adult plants exhibit an even more 
remarkable periodicity in hair formation. They commence about March 
by unfolding glabrous leaves in precisely the same way as the erect flower- 
ing shoots. Similarly, too, as the season advances, the successive leaves are 
increasingly hairy. This goes on till about midsummer (the actual time 
appears to vary, possibly with the kind of season, from about the middle of 
June to the middle of July), when the process is reversed. The new leaves 
unfolded are now decreasingly hairy , until finally glabrous leaves are once 
more produced on the approach of autumn. The actual number of hairy 
or partly hairy summer leaves formed on a given shoot appears to depend 
on its size and vigour. If a shoot is fairly large and vigorous, it may form 
only a few glabrous leaves, but a number of nearly or completely hairy 
ones. On the other hand, a small, weak shoot may produce only one 
or two slightly hairy leaves ; all the rest formed during the year being 
glabrous. 2 
To sum up, the successive leaves on an erect flowering shoot exhibit 
a continually rising curve of hairiness, the spring leaves being glabrous and 
the summer leaves hairy. On the other hand, the non-flowering shoots 
have glabrous spring and autumn leaves, and (if the plants are sufficiently 
vigorous) more or less hairy summer ones. In other words, the curve 
of hairiness rises from zero in spring-time to a maximum about mid- 
summer, once again falling to zero towards autumn. 
(d) Distribution of hairs on the partly hairy leaves . The series of 
leaves shown on PI. LXXXII and LXXXIII, selected from various non- 
flowering shoots, serves to illustrate the rules governing the localization of 
hairs on leaves which are not completely hairy. In each leaf, taken as 
a whole, it is readily seen that hairiness decreases from above downwards. 
Out of many hundreds of leaves examined, I have only met with one 
or two partial exceptions to this rule. If individual partly hairy leaflets 
are considered, it is equally obvious that the distribution takes the form 
of a marginal band of hairs (PI. LXXXII, leaves 2, 3, 5, 7, 11, &c., and 
PI. LXXXIII, leaves 1 and 2) ; with sometimes, but by no means always, 
1 Ascherson and Graebner (’00-’05), p. 438, state that the radical leaves are less hairy than the 
cauline in certain varieties of S. Ulmaria. 
2 Curious cases are sometimes met with in which a fairly vigorous shoot shows no sign of 
elongating to a flowering shoot at the usual time. It proceeds to behave like a non-flowering 
shoot, showing first increasing, then decreasing, hairiness. A few weeks later it suddenly alters its 
behaviour again ; the leaves once more exhibit increasing hairiness, and a belated flowering shoot is 
formed. 
