50 
NATURE 
[May 15, 1873 
(Figs. 1, 2), the two upper ones of which, a, 4, have no 
colouring, the two lateral petals c, d, have each one con- 
spicuous broad streak, and are furnished near the base 
with a tuft of hairs ; while the lowest, ec, has a number of 
streaks, usually either 5 or 7, and is also provided with a 
tuft of hairs near the base ; this petal is prolonged below 
into a spur, All the streaks, on both the lateral and the 
lowest petal, point exactly towards the centre of the flower 
J, where are the stamens and pistil. The stamens (Figs. 
3, 4, 5) are also five in number ; the filaments, a, are very 
short ; the anthers, 4, form a circle surrounding the pistil, 
closely applied to it, and also closely touching one 
another at their edges ; each anther has the connective, 
¢, prolonged above into an orange-coloured appendage ; 
and these also, somewhat overlapping one another, form 
a complete ring round the pistil. Two of the stamens are 
prolonged below into remarkable kneed appendages, both 
of which project down into the spur of the lower petal, 
partially filling it up. The pistil (Figs. 6, 7) consists of a 
nearly globular ovary, a, an irregularly curved style, 4, 
much narrower below, and furnished in front with a re- 
markable wedge-shaped black line, c, and of a single 
stigma, @, hooded in shape, the viscid stigmatic surface of 
which is contained in a deep cavity near its summit. In 
the open flower, this stigma (e, Fig. 3) has a most gro- 
tesque resemblance to a monkey’s or old man’s face. The 
anthers open laterally and rather within, for the discharge 
of the pollen, so that it falls naturally on the lower part 
Fic. 1.—1, Flower of Viola arvensis a,, wpper petals ; c, d, lateral petals 
¢, lower petal ; f centre of flower. 2, The petals separat-d ; c, d, lateral 
petals ; ¢, lower petal. 
of the style, which they completely invest, and it is diffi- 
cult to see how, without artificial means, any of it will 
reach the stigma; the flower is also distinctly pro- 
tandrous, the stigmatic cavity not being fully matured till 
the flower has been some time open and the pollen fully 
discharged. The “nectary,” or part specially devoted to 
the secretion of the honey, is the termination of the two 
appendages of the stamens which project into the spur of 
the corolla (indicated at 7, Figs. 3.and 5). When the 
sweet juice is collected here in sufficient quantities, it 
drops down into the bottom of the spur, to which all 
access of rain is prevented by the hairs that fringe the 
petals around the entrance of the passage to the spur. 
With regard to the fertilisation of the violets, which, as 
has been mentioned, can obviously scarcely take place 
without foreign aid, Sprengel gives a long and very full 
description of the manner in which the sweet violet is 
visited by bees and humble bees, the insertion of whose 
proboscis into the spur of the corolla, and then its with- 
drawal, will necessarily remove some of the pollen, and 
bring it into contact with the stigma either of the same 
or of a different flower. It seems hence to have been 
assumed rather than observed that the wild pansy is 
fertilised in the same manner ; although Sprengel states 
that he has not usually seen this species visited by bees, and 
Miiller’s observations* are by no means decided. My own 
* Die Befruchtung der Blumen durch Insekten und die gegenseitigen 
Anvpassungen beider : von Dr. Hermann Miller, Leipzig, 1873. 
view is that the wild pansy is fertilised chiefly, if not en- 
tirely, by very minute insects of the Thrips kind. During 
a long observation one morning this spring of a field in 
which these flowers were very abundant, I never once 
saw them visited by a humble-bee or other large species, 
and the only insect observed to frequent them was a little 
species of Thrips, and these only in small numbers, which 
I attribute to the circumstance that my only opportunity 
was the first warm sunny morning after a long period of 
cold weather, when but few insects had yet left their 
winter retreats. Sprengel indeed says that the wild pansy 
is greatly frequented by Thrips, although he believes the 
fertilisation to be effected by bees. 
If this view be correct, the markings of the flower 
furnish the insect with a most remarkable series of guide- 
posts (or, as Sprengel terms it, “Saftmaal”) to the nectar 
which serves as its food. The streakings on the lateral 
and lower petals form a sure guide, as soon as the little 
visitor reaches the flower, all converging (as shown in 
Fig. 1) to the centre of the flower and summit of the ring 
formed by the connectives of the anthers. Here even a 
minute Thrips can with difficulty force its way between 
the style and the closely adjacent ring of anthers, the 
deep orange tips of which would naturally attract it; but 
here it meets with a most curious and valuable assistance 
Fic. 2.—3, Pistil and stamens ; a, filaments; 2, anthers; c, connectives; d, 
appendages to lower stamens; e, stigma; 7, honey-glands. 4, Lower 
stamen (enlarged); 4, anther; c, connective; d, appendage 5, The 
same, seen within the spur of the corolla. 6, Pistil; a, ovary ; 4, style ; 
c, wedge-shaped streak; 2, stigma. 7, The same, seen laterally ata 
later stage. 
in the wedge-shaped streak on the front side of the style 
(as seen at cin Figs. 6 and 7), the broad upper end of 
which is distinctly visible above the anther-ring, tapering 
downwards to a sharp point near the bottom of the style, 
where the insect would be at once landed on the upper 
part of the kneed appendages, along which it has now 
simply to descend until it reaches the nectar, the object 
of its journey. The style is much narrower towards the 
base than above, and hence there is room for a con- 
siderable accumulation of pollen here, as it escapes from 
the anthers. The insect must necessarily carry away a 
considerable quantity of the pollen in its descent and 
ascent of the style ; whether for the purpose of pollenising 
the stigma of the same or of a different flower is not at 
first sight clear. The heteracmy of the flower (é¢. the 
male and female organs being mature at different periods) 
favours the idea of cross-fertilisation, which may very 
well happen from the active little Thrips visiting many — 
flowers in the course of a day. The ovules of the wild 
pansy are indeed abundantly fertilised, much more gene- 
rally, in fact, than those of the sweet violet, the mature 
capsules of which frequently result from the unopened, 
self-fertilised, “ cleistogenous” flowers, which have not, 
as far as I am aware, been observed in the pansy. 
ALFRED W. BENNETT 
