VOL. Xt-IV.J PHILOSOPHICAL TRANSACTIONS. 383 



soon after put out branches, and thus retire to a greater distance from each other. 

 It is now necessary to take notice, that when 2 of these polypi are thus formed 

 by the division of one, the one of them usually is much larger than the otherr 

 this larger one remains at the extremity of the branch where it was, but which 

 branch lengthens itself more while the other puts out a new branch, which 

 seems to proceed from tlie first. The larger of these polypi again divides itself 

 generally before the other, and every thing above described is reiterated several 

 times. Thus a principal branch is formed, provided with several lateral ones. 

 These lateral branches become principal, with regard to those which in their turn 

 seem to spring from them, when the polypi at their extremities come to divide. 

 All the polypi of a cluster do not detach themselves from it at the same time : 

 those which are nearest to the origin of the branches usually detach themselves 

 first. And every polypus so detached, goes and fixes itself elsewhere, every one 

 thus becoming at last, if not prevented, the principal of a new cluster. 



He has often kept polypi of this kind, in glasses of the size of that repre- 

 sented in fig. 7- And the first cluster he had placed in it, to observe its 

 growth and progress, continued still well provided with polypi, when there were 

 already numbers of other clusters formed in the same glass, all which owed their 

 being to those that had detached themselves from the first cluster. He had seen 

 sometimes portions of the peacock's feather in the water entirely covered with 

 these clusters : and he was well assured that all these clusters came from the first he 

 had lodged in the glass. Nay he had even carried his experiments so far as to be 

 well assured, that every polypus of a cluster, as soon as detached and fixed else- 

 where, became the principal of a new cluster. He mentioned this fact particular- 

 ly, because he should make some use of it hereafter, when he came to take notice of 

 a difference between this species of polypi, and another species mentioned aftei-wards.. 

 When a cluster is in good part stripped of its polypi, the branches are no 

 longer able to contract with the same quickness and readiness as before. Where 

 there remain but a very few polypi, none but those branches to which polypi are 

 still fixed continue to exert this power; which these also lose as soon as they are 

 stripped of their few remaining polypi, after which they show no further capa- 

 city of moving. 



From all which particulars it seems to result, that this motion in the stem, 

 and in the branches of a cluster, is entirely derived from the poljpi, which are 

 fixed on the branches. And yet an observer, attending to the appearance only 

 of this motion, can hardly help persuading himself at the first, that it is the 

 branches which draw and give motion to the polypi. 



The resemblance and the analogy which the figure of a cluster of polypi bears 

 to the figure of a plant, would induce any observer for some time to imagine that 

 the polypi, which he sees fixed to the branches of the cluster, do really proceed 



r 



