210 Mr Harmer, On Gyclostomatous Polyzoa. [Oct. 26, 



I am the more inclined to consider the last point because of 

 the views which have just been expressed by Dr Gregory 1 , who 

 reluctantly accepts the conclusions that " there are no true genefa 

 among Cyclostomata, but only certain convenient, but artificial, 

 groups of species " ; that " there is no hope in this order of ever 

 establishing divisions with the same ' absolute diagnoses as in 

 most other groups of Invertebrates " ; and that " Zoarial characters 

 are the only ones available for systematic work among the Cyclo- 

 stomata." Dr Gregory's classification is based mainly on the 

 characters of the entire colony and of the zocecia. It appears to 

 me that 'he has omitted to notice sufficiently a most important 

 adjunct to classification, namely the ovicells, whose value in classi- 

 fication has previously been insisted on by Waters 2 and myself. 



I fully admit the difficulty of distinguishing genera and species 

 among the Cyclostomata, and I by no means assert that it is 

 always possible to determine the species to which a fragment of a 

 Cyclostome, or even a single colony belongs. The ovicells, which 

 appear to me to form the surest diagnostic features, are unfortu- 

 nately not always available ; and on the assumption that they are 

 modified zooecia there was probably a period in the evolution of 

 the group during which the fertile zooecia were not to be dis- 

 tinguished externally from the others. But it appears to me that 

 in dealing with recent species at least, the study of a large series 

 of specimens, in various stages of their existence, and particularly 

 during their reproductive period, enables us to establish " absolute 

 diagnoses " which are in no way different from those which dis- 

 tinguish the members of other groups of Invertebrata. It may 

 not always be possible to apply these diagnoses in a particular 

 case, but I think that it would be going too far to assert that the 

 differences do not exist. 



In spite of the agreement of Crisia, Idmonea and Lichenopora, 

 the three genera which I have specially studied, in the fundamental 

 fact of the occurrence of embryonic fission, the details of their 

 development and particularly the structure of the ovicells and of 

 the embryophores are widely different in the three cases. This I 

 shall hope to show, in greater detail, in a future paper. It is true 

 that these are extreme cases ; but the differences are not confined 

 to extreme cases, as the following instances will show. 



I have on a previous occasion attempted to show that the British 

 species of Crisia can be distinguished from one another by their 

 ovicells 3 . Another genus in which the diagnostic value of the 



1 J. W. Gregory, "Catalogue of the Fossil Bryozoa in the Department of Geology, 

 British Museum." 8vo. 1896, p. 21. 



2 Mr Waters has returned to this point, in a review of Dr Gregory's Catalogue 

 published in the November number of Natural Science (Vol. ix. 1896, p. 334). 



3 Quart. J. Micr. Sci. xxxn. 1891, p. 127. 



