168 Transactions of the Boyal Society of South Africa. 
creature with a broad head and strong mandibles, the three pairs of long 
legs terminating each in three tarsal claws. 
In the case of Meloe and certain kinds of Gantharis feeding on honey 
and possibly bee-bread, the Triungulin or primary larva climbs the stem of 
plants and awaits there the arrival of bees or other flower-frequenting 
insects, hooks itself on their hairs, and, if its carrier happens to be a bee, 
it is conveyed into the nest. It penetrates a cell during the oviposition, 
or possibly after the storage of the cell is already completed, and devours 
first the egg of the rightful owner and turns into the second larva which 
assumes quite a different aspect, having on that account the name 
caraboid-stage given it ; it is soft, and eventually assumes the third stage, 
that of pseudo-pupa, but still within the second larval skin ; it is then 
mostly motionless ; in the next stage this pseudo-pupa turns to a larval 
form closely resembling that of the second instar, but also without shed- 
ding its skin, being thus wrapped into two ; in this envelope it changes 
to the true pupa, and lastly appears as the perfect insect. It has then 
undergone six transformations. 
According to Eiley, these post-embryonic developments are a little 
more complicated in the cases of some North American species of Gantharis 
(Epicauta), the young of which feed on locust eggs. The active triungulin 
does not cling to the hairs of bees, but sets about finding the egg-pods of 
locusts. Should two or more triungulins penetrate the same egg-pod, 
they fight until only one is left. The instars are then as follows (ap. Sharp) : 
(1) Triungulin larva-moult ; (2) Caraboid larva-moult ; (3) Scarabseoid 
larva-moult ; (4) Scarabseoid larva ultimate stage-moult ; (5) Coarctate 
larva-moult ; (6) Scolytoid larva-moult ; (7) pupa-moult ; (8) perfect 
insect. It may, however, be added that Eiley's three and four instars do 
not differ much from the sixth. 
Sitaris deposit their eggs at the mouth of the galleries of burrowing 
Hymenoptera. In one species observed by Fabre the triungulins pass the 
winter there, motionless, under the shell of the egg, before clinging to the 
hairs of the passing Anthophora that are to introduce them into the 
galleries of the nest. But in other species the triungulin fastens itself at 
once to the GoUetes, Halictus, or Andrena bees entering the nests. 
The different stages in this hypermetamorphosis vary in duration ; the 
habits of the larvae of different species and genera are also presumably not 
the same.'" Certain species are certainly produced during a season, others 
are not. It seems from all observations that it is in the pseudo-pupal 
state that the stage is the longest, lasting occasionally over a year. In 
Gantharis vesicatoria Beauregard finds that the triungulin, unlike in that 
respect that of Meloe and Sitaris, does not devour the hymenopterous egg 
laid in the cell, but feeds on the honey at once, and that the second larva 
* See p. 290. 
