On the Species of the Genus Primula in Essex. 161
with the Primrose, although I am quite unable to account
for this difference. Nevertheless the occasions on which the
1. outnumbered the s. were more than twice as numerous
as those in which the reverse was the case. If we may
judge from Obs. 16, in which I gathered 65 of the finest
umbels I could see, what was said of the last species, as to the
long-styled umbels tending to be the finest, does not appear
to apply to this—indeed the exact opposite is the case, no less
than 42 of these being short, or in the proportion of 100
s. to 55 1.
Mr. Darwin has stated that the long-styled plants of
the Cowslip and P. sinensis tend to flower first. From this
it would seem probable that such was also the case with the
other English species of Primula, but I have been unable to
observe this; and the only evidence which Mr. Darwin him-
self brings forward in support of his statement is that he had
twelve Cowslip-plants of each form grown under similar con-
ditions, and at one time seven long-styled ones were in flower
against one short-styled. My first four Cowslip observations
are good tests of this statement, as they were made in places
where only a few of the very earliest plants had flowered;
yet in the second observation the numbers of the two forms
were exactly equal; and in the 1st, 8rd, and 4th the result
was just as usual—the 1. slightly outnumbering the s.
Observations on the Oxlip tend to show that, in this species
at least, the opposite of Mr. Darwin's statement is the case;
for in my first observation, made as early as March 12th, I
managed with great trouble to obtain 231 umbels with
flowers only just opened, and in these the s. outnumbered
the 1.; but three days later in another locality the 1. had their
usual majority.
Supplementary Remarks on Primula farinosa, &c.
Table IV. shows that, of 804 plants of P. farinosa examined
on nine different occasions in the Upper Engadine between
July 5th and August 23rd, 1882, 365 were short-, 436 long-,
and 3 equal-styled, thus giving the 1. a clear majority of 71.
On a single one only of these occasions did the s. outnumber
the 1., and then only by 4 out of 14, the numbers being too