MB. A. W. BENNETT ON PAKNASSIA PALUSTRTS. 27 



r ' ~ 



- ,v _ 



' r I 



they only begin to display themselves and to expose their papil- 

 lose tongues at the moment when the emission is accomplisiied " 

 (vol. i. p. 324). The successive lengthening of the filaments was 

 observed so long ago as by Sir James Edward Smith; and the 

 manner in which this takes place is very remarkable. The 

 increase, to the extent of at least three or four times their original 

 length, must be accomplished in an incredibly short space of time ; 

 the adhesion to the ovary is so strong during the whole of this 



L 



time that they cannot be bent back without breaking them ; but 

 as soon as the pollen is discharged, they retire to a horizontal 



^ _ - J J V 



position between the petals, and the anther falls. My own ob- 

 servation does not, however, confirm Vaucher's statement that 

 the lengthening takes place alternately ; I have frequently noticed 

 contiguous stamens to follow each other. It will be observed 

 that the movement of the stamens in JParnassia presents but little 

 resemblance to the "approach of the stamens to the pistil in 

 pairs/' which is described as taking place in certain species of 

 ^acoifraga ; nor does it appear in this latter genus to be accom- 

 panied by the simultaneous lengthening of the fiiamentj which 



serves an important physiological function. Together with this 

 elongation of the filament, and previously to the discharge of the 

 pollen, a singular contraction of the anther takes place; and 

 I have no hesitation in concluding that the arrangement above 

 described is one of the most remarkable provisions of nature yet 

 observed for insuring cross-fertilization ; for not only does the 

 anther place itself, at the time of the ripening of the pollen, with 

 its back on the very apex of the pistil, so as completely to close 

 the approach to the ovary, but, as if to make assurance doubly 

 sure, the stigmata are not developed until the whole of the 

 anthers have successively performed this movement and dis- 

 charged their pollen. The object of the glandular nectaries is 

 now clearly seen, and is not, as Vaucher imagined, the return of 

 the pollen to its own stigma, but to enable insects to carry it 

 away to other flowers in which the stigmata are already expanded. 

 I spent a considerable portion of oi^Q of those rainy mornings 

 which in Scotland bring forth such countless clouds of insects, in 

 keeping watch over a field as thickly studded with Farnassia as 

 an English hedge-bank with primroses, and scarcely noticed a 

 single flower in which several insects were not regaling themselves 

 on the nectariferous glands — belonging to several species, but 

 mostly a long-legged dipterous fellow, whose long thighs, straddling 



