POLYZOA OF WATERWORKS. 445 



that the formation of the hihernaculum is inaugurated by a con- 

 traction of the endocyst or body-wall from the original ectocyst 

 of the zocecinm, in the distal part at least, and by the deposition 

 of a new cuticle over that part of the hibernaculum which lies 

 freely inside the old ectocyst. In some cases, as in fig. 10, 

 the hibernaculum has a wall of its own which is distinct 

 throughout from the original ectocyst. In the specimen shown 

 in fig. 2 a single zooecium has developed two hibernacula., the 

 first of which is attached to the pa.rent-zocecium in the position 

 of a terminal bud, while the second, or distal hibernaculum, lies 

 freely in the cavity and has no base of attachment. It can 

 hardly be doubted that the hibernacula represented in figs. 3, 9, 

 and 10 would have assumed the appearance shown by most of the 

 hibernacula in this material by the loss of the empty caps formed 

 by the zooecia within which they have severally been developed. 

 I regard it, theiefore, as pi'obable that the hibernaculum owes 

 its terminal position to the loss of the distal part of its own 

 zocecium, which, as shown in fig. 10, is not necessarily the 

 terminal individual of a branch. 



There are, however, certain appearances, shown in figs. 6-9, 

 that are in need of explanation. In these cases most of the 

 ectocA'st of the zocecium which precedes the definitive hiber- 

 naculum has split into two valves, in the manner characteristic 

 of a germinating hibernaculum. Like the other zooecia of this 

 material, these zooecia are destitute of cellular contents ; the only 

 living tissues left in the colonies being those which constitute the 

 hibernacula. It Is important to consider whether the bivalve 

 arrangement indicates that a process of germination has actually 

 taken place. In view of the close resemblance between these 

 zooecia and a hibernaculum germinating after the end of the 

 winter and arrived at the stage of having completed the formation 

 of the first regenerated zooecium, it seems legitimate to assume 

 that germination of a hibernaciilum has occurred in spite of the 

 fact that the material was collected in the autumn. It would 

 appear that a new zooecium has been completely formed, since 

 the remains of its tubular orifice are present. But the develop- 

 ment of a new colony has stopped short with the formation of a 

 single zocecium, and a second process of hibernaculum-develop- 

 ment has taken place, resulting in the definitive winter-buds. 

 The living material of the regenerated zooecium has passed into 

 tlie hibernaculum, or hibernacula, which have been formed from 

 it, though there is no evidence to show whether the transference 

 of the material took place en masse or by the more gradual process 

 by which a bud is ordinarily supplied with its living contents. 

 It is perhaps not impossible that no polypide was formed by the 

 regenerated zooecium, but that on the splitting of the hiber- 

 naculum which jjreceded it the contents of that structure passed 

 bodily into the newly-formed hibernaculum. Some such ti^ans- 

 ference of cellular material may be indicated by fig. 10, Avhere 

 each of the hibernacula is seen to traverse the septum which 



