446 DR. S. F. IIARMER ON THE 



occurs between two zocecia. In fig. 8 the definitive hibernaculum 

 may be proved by measurement to be of exactly the same length 

 as the hibernaculum which is indicated by the split valves in the 

 wall of the preceding zooecium. The material of the first hiber- 

 naculum would thus just fill the cavity of the second one. In 

 fig. 9, however, the second hibernaculum is distinctly smaller 

 than the first one ; while in fig. 7, what is assumed to have been 

 the first liibernaculum has had to provide the material for two 

 new hibernacula. Further observations on material collected at 

 a more favourable season will be required to settle these points. 



The great majority of the zooecia in a PcdadiceUa-colony give 

 rise to a pair of lateral buds, situated opposite one another and in 

 a position which has a definite relation to the orifice of the parent- 

 zooecium. It may be significant that in all the cases shown in 

 figs. 6-9 there is some abnormality in this respect. In figs. 6 

 and 9 only one of the lateral zocBcia is indicated ; while in figs. 7 

 and 8, where two are present, they are not opposite one another, 

 and in fig. 7 one of them is much nearer the proximal end of the 

 parent-zooecium than is normally the case. I have observed 

 similar ii^regularities in other zocecia of the same kind which have 

 not been figured ; and these cases perhaps support the view that 

 germination of the hibernacula first formed has taken place, since 

 it is well known that in young Polyzoon colonies the zooecia first 

 developed frequently fail to give rise to the full number of buds 

 which are formed by the zooecia in an actively growing branch. 



If the explanation above suggested is correct, the germination 

 of the first set of hibernacula may perhaps have been induced by 

 tlie occurrence of warm weather succeeding a cold period, which 

 induced the formation of the first set. But the explanation in 

 question is not the only one which can be imagined ; and it is 

 not inconceivable that as the time of hibernaculum-formation 

 approaches, some of the zooecia make an abortive attempt to 

 develop hibernacula, and that the bivalve arrangement seen in 

 figs. 6-9 implies an unsuccessful efibrt of this nature. 



In the majority of the zooecia of this material the definitive 

 hibernacula have been produced at the first attempt, and the 

 zooecia which respectively precede them show no appearance of a 

 bivalved arrangement. But in all cases the zooecia are empty of 

 cellular contents, and there has obviously been a centrifugal 

 passage of the living material of the colonies into the hibernacula, 

 which are finally left as the outermost members of the series. 



It remains to be considered whether these observations have 

 any bearing on the question of the homology, if any corres- 

 pondence really exists, between the hibernacula of Paludicella 

 and the statoblasts of Phyla ctolfemata. It has frequently been 

 maintained that statoblasts have been evolved from structures 

 more or less resembling the hibernacula of Paludicella^ which on 

 this view is regarded as related to the ancestral form from which 

 the Phylactoljemata have been derived. This argument has been 

 specially developed by Kraepelin (87, pp. 163, 167 ; 92, p. 62), 



