MR. T. Н. CORRY ON ASCLEPIAS CORNUTI. 191 
the insect, firmly fastened to its foot, and follows the course of its leg up the chamber*. 
Both of the pollinia may occasionally be left, but more usually one only is so. If one 
is withdrawn I should imagine that it turns its larger diameter more obliquely, and so 
is enabled to pass through the contracting channel. ‘The leg which bears the corpusculum 
carries with it, attached to it, considerable portions of either one or both of the broken 
and detached “appendages,” since the rupture occurs either where the pollinium joins 
the appendage, though this is rare, or more commonly at the flexure, a small portion, 
generally only that below the flexure, remaining attached to the pollinium. When the leg | 
so provided has pursued its way upwards and arrived at the superior end of the alar 
chamber, а good outward pull on the part of the insect is required to extricate it. It 
thus happens that the broken end of an “appendage” becomes firmly caught in the 
wedge-shaped fissure of the corpuseulum which is seated there. This new corpusculum, 
with the two pollinia appended to it, is then extracted and carried off attached to the 
former corpusculum by means of the broken **appendage." A fresh corpusculum, with 
its appendages and pollinia, is thus carried away by the insect each time it completes the 
pollination ofa flower, if it has not been already removed ; and from the foregoing account 
it will be clear that the insect alternately gets pollinia attached to itself and again loses 
them. Each corpusculum, then, is placed at the superior end of an alar fissure, not 
alone because it is very small and unlikely otherwise to be touched by the insect, as 
Sprengel thought (loc. cit.), but in order that insertion of the pollinia may be accomplished 
by the same mechanism which serves for their extraction, and may, sometimes at least, 
immediately precede the latter part of the operation. 
Most curious and remarkable combinations of corpuscula-bearing pollinia with one 
another are occasionally to be met with, and were first figured by Brown +, although he 
says nothing whatever about the method by which they were produced. These combina- 
tions are either attached to unremoved but displaced corpuscula, or to the claws of the 
insects which effect the pollination of these flowers, and which present a most singular 
appearance when equipped with them. Тһе simplest case of such a combination which 
can occur, viz. two corpuscula and three pollinia, has been already noticed, together with 
the mode in which it is formed. This combination being inserted into the alar fissure, as 
the single corpuseulum was previously, one of the two lower pollinia is left detached in 
the lower part of the chamber, whilst its appendage, becoming caught at the upper part 
in the fissure of the corpusculum which lies there, carries it away, forming a com- 
bination of three corpuscula and four pollinia. This may be repeated again and again 
with inevitably the same result : one of the pair of pollinia last extracted is left in the ајаг 
chamber, while a new corpusculum, with its two pollinia, is each time withdrawn, and so 
a row is formed. These corpuscular combinations are, however, frequently more or less 
+ From some observations of Dr. Hermann Miiller’s, it is clear that a strong insect is able to insert pollinia and 
then escape with the corpusculum, without, however, extracting another also. This it must do by a powerful oblique 
pull, so that the foot is forcibly withdrawn before it can travel the whole length of the alar chamber. Müller found 
pollinia inserted into the alar chambers of flowers from which no corpuscula were withdrawn ; and this has also been 
confirmed by Mr. Weale’s researches (oc. cit. p. 54) оп Gomphocarpus fruticosus and G. physocarpus. 
+ Vide plate xxxii. figs. 4-6, Atlas of plates to Misc. Bot. Works. 
262 
