REPRODUCTIVE ORGANS IN THE PRIMULACER. 115 
styled form there is a converse arrangement of these organs 
observed, the stamens corresponding, or nearly, in length with 
the pistils of the long-styled form, and the pistils with the 
stamens of the latter form. The stigmas of the two forms differ 
little in shape or size, but that of the short-styled form is 
evidently rougher. The pollen-grains are also of a similar 
(bluntly triangular) shape in both forms, those of the short- 
styled being the larger. Besides these two forms, however, 
characterized as we have seen by a relative inequality in the 
length of the stamens and pistils, it is not at all uncommon to 
find individual plants with these organs of an equal length, and 
reaching the mouth of the corolla-tube. From an examination 
of specimens, however, from various localities, I have no doubt, 
from the relatively small size of the pollen-grains as compared 
with those of the normal long-styled form, that these relations 
are due to an abnormal development of the stamens. This view 
is furthermore supported by the functional performances of 
these organs, as I find, from a few experiments on an equal- 
stamened and -styled plant in the Botanic Gardens of Edin- 
burgh, that less seed results from its union with the long-styled 
than with those of the short-styled. The non-dimorphic form of 
P. Auricula (vide Table II.) affords an analogous illustration in so 
far as concerns its functional relations with the long- and short- 
styled forms of the species. But there is one important differ- 
ence between the two cases, namely, that the non-dimorphie 
form of P. Auricula is perfectly fertile with own-pollen, whereas 
the like form of P. farinosa is very imperfectly fertile when thus 
treated! indeed much less so than either homomorphic union. 
This is seen by referring to the table beneath, where we find that 
the united homomorphie unions yield more seed, in about the 
proportion of two to one, than those of the equal-stamened and 
-styled form by own pollen. 
In the following table I have given, first, the unions with the 
equal-stamened and -styled form by own pollen, and secondly, the 
reciprocal unions of the long- and short-styled forms. 
