72 N. AnnandaLE : The Fauna of Brackish Po)ids. [VOL. I, 



vicissitudes to which they are exposed in the ponds at Port Canning, 

 it is very improbable that any of the individuals now living in 

 these ponds have survived for so long a period, while the presence 

 of numerous young in the ponds and of ripe gonads in the adults 

 proves that we are dealing with a race and not merely a collection 

 of infertile individuals. The modifications are undoubtedl}^ less 

 marked in the young than they are in their parents, between which 

 and the typical form the young are intermediate. Tliis is true as 

 regards biological as well as structural characters. The youngest 

 individuals of the t^^pical form I have seen (measuring about 4 mm. 

 in height) have had a considerably shorter column than examples 

 of the isolated race with disks of a smaller diameter. 



Variation has been little studied in the Actinians, which do 

 not make satisfactor}^ specimens either for the museum or the 

 laboratory ; but the stony corals, in which the skeleton preserves 

 in mau}^ respects a complete diagram of the living tissues, prove 

 how variable certain genera and species of Zoantharia can be (for 

 example see Bernard on Pontes in the Catalogue of the Madrepo- 

 rarian Corals in the British Museum, vol. v, 1905). I doubt whe- 

 ther Gosse was so far from the truth as later systematists believe 

 him to have been when he laid stress on the importance of the study 

 of the living organism in the case of the Actinians. It is worthy 

 of note that, at any rate as regards the Sagartiidse, the descriptions 

 of genera have recently shown a tendency to become more rather 

 than less indefinite. Compare, for example, Hertwig's (4) defi- 

 nition of Sagartia, published in 1882, with Haddon's (8), published 

 in 1898, or with McMurrich's (10), published in 1905, having regard 

 to the fact that these authors are in substantial agreement as to the 

 species which should be included in the genus. As the three diag- 

 noses are short, they may be quoted in full : — 



" Sagartiidce with smooth walls and numerous powerful ten- 

 tacles arranged in several rows ; with circular oral disk ; 

 without anatomically perceptible cinclides." (Hertwig, 

 1882.) 



" SagartiincB with a smooth body- wall, or with small verrucae 

 in the upper portion of the column ; moderately long ten- 

 tacles in several cycles around the margin of the oral 

 disk, which is not greatly expanded." (Haddon, 1898.) 



' ' Sagartiince with the column smooth or provided with verrucae 

 in its upper portion ; cinclides more or less scattered ; 

 acrorhagi wanting ; margin not lobed." (McMurrich, 1905.) 



The diagnoses of the family and sub-family given by these 

 authors are still more diverse, but the point I wish to bring out 

 is the way in which various descriptions illustrate the necessity 

 felt by recent authorities for broadening the diagnoses of Actinian 

 genera. 



