REPRODUCTIVE ORGANS IN THE PRIMULACEE. 79 
anthers project far out of the flower, and correspond in height 
with the projecting stigma of the other form, as does the short 
pistil with the short stamens of the long-styled form. But the 
most remarkable difference is in the size of the pollen-grains : 
those from the short-styled flowers, when distended with water, 
are 1455; of an inch in diameter ; those from the long-styled (and 
therefore from the shorter stamens) are ;;%; 7 in diameter. 
Spheres differing in the proportion of 14 to 9 in diameter pre- 
sented under the microscope a most remarkable contrast. The 
contents of the larger pollen-grains seemed to be coarser-grained 
and of a browner colour.” 
To these structural observations of Mr. Darwin I may add a 
few experiments which I have lately had an opportunity of making 
upon a * short-styled " plant of this species, which, though worth 
little, afford more than a negative presumption that the above- 
described morphological peculiarities are also, as in the genus 
Primula, connected with remarkable functional relations. My 
experiments were limited to two racemes. In one I fertilized 
twelve flowers with own pollen; these yielded six capsules, five 
of which were good, and contained in all thirty-seven seeds. The 
other raceme I fertilized with pollen sent me of a long-styled form. 
The results from eight flowers thus fertilized were, jive good cap- 
sules, and one hundred and fifty-four seeds. Thus, in the “ homo- 
morphic unions"—or union with own-form pollen—the average 
production of seeds per capsule is about seven, while in the “ hete- 
romorphic union "—or that resulting from the union of the two 
forms—we find the average of seeds per capsule increased to thirty 
and upwards. 
Before passing from this genus, I may state that the H. inflata 
of North America does not present structurally dimorphie cha- 
racteristics, however it may be as respects function. In speci- 
mens which I have examined, I found the style very short, and 
reaching the base of the anthers; the pollen-grains very similar 
as respects size, i.e. judging from dried specimens, to those of 
the “long-styled” form of H. palustris. 
2. The genus Primula has, along with a great majority of 
truly “ dimorphic” species, others in which I have observed the 
one form only; also a few with stamens and pistils of an equal 
length—* aon-dimorphic." In the following list I have enume- 
rated all the species of Primula which I have had an opportunity 
of examining, and arranged them in accordance with my obser- 
vations on the structure of their sexual organs. The first list 
