Incarvillea Delavayi , Franch. 69 
very minute papillations. The exine had long slits very much like those of 
Thunbergia (Fig. 3, A, B, C, D), but while in this genus the slits form one or 
two continuous lines, in Incarvillea the grooves only occasionally join 
together, so that we never get the appearance seen in Fig. 295 — V, VII, and 
IV — of the ‘ Outlines’ (Goebel) (6), the closest approximation being shown in 
my Fig. 3, c, where a few slits will be found to join, with the result that 
a portion of the outer wall opens out after the inner wall has been 
dissolved by sulphuric acid. An opportunity presented itself of examining 
a few withered flowers of Incarvillea grandiflora , in which plant the floral 
mechanism is quite similar to our species, the difference being in details, 
such as the longer flower stalk and the slightly different colour and marking. 
The pollen of I. grandiflora proved to be similar to that of /. Delavayi , but 
there were fewer slits and a very few small adult grains were to be observed 
amongst them (Fig. 3, G, H,"K, L) ; the meaning of this heterospory is not 
clear to me. As there were two lengths of filament, pollen-grains from the 
anthers at different levels were examined in both species and closely com- 
pared as regards size and markings, especially as some such differences 
might have been expected, and in Torenia , for example, one pair of anther- 
lobes bears dry, the other pair moist, pollen ; no differences, however, 
of any kind could be noticed in the case of /. Delavayi . Whether the 
.different sized grains are commonly found in D. grandiflora or whether they 
are confined to the longer or shorter stamens, I am unable to state. In 
Torenia (2, 10), it will be remembered, only the dry pollen is immediately 
effective in pollination ; the moist pollen having to be dried before it will 
bring about fertilization. In Lagerstroemia indica ( 4 , 7 ) there are also two 
kinds of stamens — conspicuous ones, with moist pollen, visited by the insects 
and used as food, and inconspicuous ones containing dry pollen, used in 
pollination. Seemingly both kinds of pollen can in this case cause fertiliza- 
tion without any preliminary treatment. In Incarvillea the moist pollen 
need not be dried before it is effective. On exposure to air the pollen- 
grains immediately begin to lose water, get slightly smaller, and change 
shape from the spherical; the portions' of exine between adjoining slits 
heaping themselves up (Fig. 3, E) (cf. figures of Thunbergia ( 5 )), the water 
loss probably taking place mostly through the intine and the lips of the slit 
then approximating, and further loss causes a hinge-like movement along 
the crack. This approximation of the sides must considerably reduce the 
further loss by evaporation. Such drying would probably take place 
on the back of the insect carrying the pollen from one flower to another, 
though it is questionable whether the slight irregularity of surface so 
brought about is of any great importance for further fixation to the creature. 
The presence of slits and the high osmotic pressure of the contents, the 
evidence for the existence of which will be described below, I regard as 
adaptations to prevent drying. 
