Mr Brindley , The procession of Gnetliocampa pinivora. 101 
on the whole continuing to move forward in the original direction, 
but now in two parallel broken chains. By degrees the larvae 
of each chain closed up and in 10 minutes there were two solid 
chains marching parallel. 
(2) Breaking up a long chain by taking out several contiguous 
larvae here and there was tried twice, both times with the result 
that those left in place broke their formation and came to form 
a confused mass by circling round each other, the larvae removed 
soon mingling with the rest. Eventually from this mass an 
approach to a new chain was evolved, though when last seen 
there were still stragglers. 
(3) Placing 4 larvae head to tail in as nearly a closed circle 
as possible resulted in their joining up and going round in a 
closed circle several times, a proceeding soon relinquished for a 
straight procession. 
(4) Placing a single larva a little ahead of a short chain had 
the sequel of the leader of the chain catching up the isolated 
larva, which was accepted as primite with practically no hesitation. 
In making these observations it was not possible to decide 
whether vision or the thread secreted by a predecessor was the 
more useful guide wffien “joining up” occurred. It was indeed 
often difficult or impossible to distinguish the fine thread on the 
glistening sand. The different results of making small breaks or 
large breaks in the normal procession, as in (1) and (2) above, are 
recorded for what they may be worth, for far too few observations 
were made to justify exact conclusions. Fabre observed that pro- 
cessions sometimes voluntarily broke up into circulating masses, and 
in such cases fatigue after long marching, darkness, and nocturnal 
cold seemed to him possible factors. At any rate the circulating 
mass is the normal formation in the nest and a procession belated 
away from home, as in some of Fabre’s cases, might perhaps be 
expected to adopt the nest formation at nightfall. But by what 
process does the mass become a procession ? It seems quite 
clear that any larva may act as a primite, and thus the doubt 
arises as to whether the primite does, in a strict sense, lead the 
procession. It seems not unreasonable to suppose that as 
all the larvae must be influenced at the same time by the same 
stimulus, hunger, sunlight, desire to pupate and so on, they all 
respond fairly equally, and thus it may be suggested that the 
whole chain proceeds as one individual, as seems to be the case 
in a chain of salps or of polychaetes freed from the budding 
individual, though the organic union between these is of 
course absent in Gnethocampa. But in one particular case the 
primite may perhaps take a real initiative, i.e., when burrow- 
ing is performed. We observed this process in the case of 
one of the daughter chains formed in (1) above. The spot 
