160 On the Species of the Genus Primula in Essex. 
were 1. Obs. Nos. 23, 26, 29, 30, 31, 32, and 34 all point 
more or less clearly in tbe same direction. In opposition, 
however, to these examples I wish to refer to Obs. 9, in 
which (the wood being fully grown up) no stimulation could 
have come into play, and to Obs. 18, in which the flowers 
were noted as being small and going off; yet in these cases 
the s. stood to the 1. in the proportions respectively of 47 and 
51 to 100. Nevertheless it is certain that, in this species at 
least, the long-styled plants have, when stimulated, a far 
greater tendency than the short-styled ones to develop large 
umbels—often abnormally so,—and this observation may 
hereafter be found to be of some importance. 
Turning now to Table III. we see that, of 4276 Cowslip 
plants examined on twenty-nine different occasions in various 
localities in Essex between March 29tli, 1882, and April 24th, 
1883, 2125 were short-, 2142 long-, and about 10 equal-styled, 
thus giving the 1. a majority of 17. On eight of these 
occasions, however, the s. had an aggregate majority of 141 
out of 1650 plants, the relative proportion between the two 
forms ranging in these cases between 233 and 110 as com¬ 
pared with 100 ; whilst on the remaining nineteen occasions 
(two being exactly equal) the 1. had an aggregate majority of 
158 out of 2626 plants, the relative proportions of the two 
forms ranging in these cases between 54 and 99 (omitting 
Obs. 18 as unreliable) as compared with 100. On one 
occasion (No. 18) when the flowers were withering, the 1. 
actually outnumbered the s. in the proportion of 100 to 19, 
but the plants examined were too few in number to afford 
reliable results. I only give such observations that there 
may not be the slightest suspicion of my having “cooked” 
my figures by suppressing anything. On two occasions 
(Nos. 2 and 24) the two forms were exactly equal in number. 
Taking, however, the whole of the twenty-nine cases together 
we shall see that the average proportion in which the s. stand 
to the 1. is as 99 - 2 compared with 100. The foregoing remarks 
show that in the Cowslip the relative proportion by which 
the 1. outnumber the s. is not nearly so great nor so 
regular as in the case of the Oxlip ; and is rather less than 
