F. B. Browne 355 
A second moult takes place before the final larval stage is reached but 
no further change takes place in the number of spiracles. 
The full-grown larva attains a length of about T7 to 1*9 mm. and is usually 
smooth, the constrictions between the segments having disappeared during 
growth. 
It has been said that the larva feeds without puncturing its host but this 
does not seem to be strictly accurate. The larva possesses a minute pair of 
mandibles which, in the first stage, are more or less triangular in shape, the 
free angle being sharply pointed; but in larger larvae this point is greatly 
lengthened (Plate XXVI, figs. 3 and 4). I imagine that these points are driven 
in subcutaneously and serve to hold the mouth of the parasite to the host 
and at the same time allow the “blood ’ of the host to pass out. The minuteness 
of the punctures can be gathered from the fact that, on the removal of a feeding 
larva the spot at which it has been feeding is not recognisable, even under 
high magnification. On the other hand if a large number of feeding larvae are 
gently removed from a host the surface of the latter appears as if it were gently 
perspiring, though of course this might be due to exudation of fluid from the 
mouths of the larvae at the moment of removal. 
The larva is full grown, under the most favourable conditions of food and 
temperature, in eight or nine days, and only then does it begin to produce 
excrement. This habit of not producing excrement until feeding has ceased 
seems to occur in a large number of Hymenopterous larvae though there are 
others, such as Osmia, Megachile , Sapyga, etc., which begin to produce faecal 
pellets when about half grown. 
Shortly after completing this cleansing process the Melittohia larvae become 
pupae. 
It is interesting to note that the larvae take the colour of the host upon 
which they feed, as do other “blood'’ sucking Hymenoptera, e.g. Odynerus, 
Chrysis, etc. By feeding the Melittobia larvae upon the salmon-red coloured 
larvae of a small Crabro sp., the larvae became salmon-red. Fed on larvae of 
Odynerus spinipes the larvae are yellow, while if fed upon larvae of 0. antilope 
they are creamy white and this colour depends upon the blood colour of the 
host, which in turn depends upon the blood colour of the beetle- or moth- 
larva or spider upon which it fed. By feeding larvae for a few days on Crabro 
and then transferring them to Odynerus , the colour of the parasite changes 
from salmon-red to yellow or cream. 
As soon as the larva begins to cast out excrement, the colour begins to 
disappear and all Melittobia larvae, however they have been fed, ultimately 
become a pale transparent creamy-white colour. 
(d) The pupa. 
As soon as the pupa appears it is possible to decide the sex of the individual 
since, in the female pupa, the compound eyes are visible, though at first 
without colour, while no eyes are at first visible in the male pupa. In the 
