No. 403.] ECITON SUMICHRASTI NORTON. 569 
d'une heure ou plus (parfois moins) pour arriver à l'ordre, à réunir leur 
nymphes et surtout à reconnaître les environs, les Eciton se concertent, sans 
perdre une minute. En cinq minutes, ils ont formé des files distinctes de 9 
qui ne s'écartent pas les unes des autres, portent en partie les larves et les 
nymphes, et marchent en ligne assez précise, tatant le terrain de leurs an- 
tennes, explorant tous les trous et fentes jusqu'à ce qu'elles trouvent une 
cavité à leur convenance. Alors le déménagement s'exécute avec un ordre 
et une rapidité étonnants. Les 9 se suivent et se comprennent, comme par 
mot d'ordre, et en bien peu de temps tout est en lieu sûr. Pas question de 
transport mutuel; c'est en plus précis le systéme des Tapinoma. Le fait 
est remarquable chez une fourmi aveugle, qui n'a absolument que ses an- 
tennes pour se diriger. 
The colonies of Æ. sumichrasti are very much larger than 
those of E. omnivorum, and consist of thousands of individuals. 
They are found under large stones, often occupying cavities 
which they seem to dig themselves. They assemble in these 
cavities in compact masses, or swarms, which may be as large 
as a man's fist. These masses present a peculiar appearance, 
owing to the interlacing of the innumerable long, dark-red legs 
and antennz, interspersed with the shining, lighter red abdo- 
mens. If such a mass be hastily scooped up and placed in a 
glass jar, the method of swarm formation may be watched at 
one's leisure. If the jar be held perpendicularly after being 
violently shaken till all the ants fall to the bottom, they will at 
once proceed to march up the sides of the jar in regular files. 
A few ants will soon come to rest, clinging to the glass, and 
will form the nucleus of a swarm which grows rapidly by addi- 
tions from the moving columns. The ants are positively geo- 
tropic and therefore tend to hang head downwards when they 
cluster on one another. Such a swarm on the side of a jar is 
usually spindle-shaped, with perpendicular long axis. Fre- 
quently the swarm grows too large to be held in place by the 
claws of the individuals next to the glass, and the whole mass 
falls writhing to the bottom of the jar. The ants at once file 
up the walls of the jar and repeat the process. Occasionally 
the ants arriving at the top of the jar first will form the start- 
ing point for chains or festoons of ants, depending sometimes 
to a length of six inches in the form of a cord about a centi- 
meter in diameter. In forming these chains, which remind 
