LAW OF ASSOCIATION IN THE ANIMAL KINGDOM. 35 
casting themselves free, and terminating their existence in the 
colony from which they were horn. 
In one whole group of Polypes these Medusae, or their 
equivalents, associate themselves with the reproductive in- 
dividuals to form a new unit, a sort of small distinct colony, 
which may also be taken for a peculiar organ, the composition 
of which presents the most curious analogies with that of a 
flower. This reproductive apparatus has its separate chamber, 
the gonangium . 
A step further, and we shall see these Medusae, the indi- 
viduality of which is, if possible, still more strongly marked 
than that of the Polypes, themselves descend to the rank of 
organs in more complex colonies. 
All the colonies of Hydroids do not live attached to 
submarine objects. There are some which lead a vagabond 
existence. They have often, and not without reason, been 
taken for simple animals analogous to the Medusae, and desig- 
nated by the name of Siphonophora, which has remained 
attached to them. They sometimes attain a large size, and the 
variety and number, I may say the profusion of the parts 
of which they are composed, together with the brilliancy of 
their colours and the incomparable beauty of their forms, have 
always been a source of profound admiration both for the 
naturalist and the sailor. How each of these parts is the equi- 
valent of a Hydra or of a Medusa. 
In an Agalma , as in the Hydractiniae, we find nutritive 
individuals furnished with a long tentacle, the mere contact of 
which produces an intense burning pain, — a kind o {fishing-line, 
capable, in the larger species of Siphonophora, of capturing even 
fishes. Side by side with these nutritive individuals, we find 
others destitute of mouths, which are nothing but the repro- 
ductive individuals, in the vicinity of which are the sexual 
individuals, in form very like the Medusae. All these indi- 
viduals are fixed upon a common axis, floating and undulating 
in the water, sustained by a sort of aeriferous bell, which forms 
its superior extremity ; two series of sterile Medusae beneath 
this bell simulate a set of rowers, to which the entire colony 
abandons the duty of moving it about. 
In every respect these various parts are so like the Hydrae 
and Medusae, that we cannot for a moment refuse to accord to 
them the character of individuals ; the Agalma and the other 
Siphonophora are, therefore, true societies or colonies. But 
then most of the individuals cannot, without risk of death, 
be separated from their companions : and, indeed, in certain 
cases, all co-ordinate their movements to enable the colony to 
accomplish certain actions. For example, the Portuguese mcn- 
of-war (. Physalice ) are often seen to change their course, and 
