530 F. W. GAMBLE AND J. H. ASHWORTH. 



and partly of Branchiomaldane, a form described origi- 

 nally by Langerhans from the Canary Islands. 



These examples suffice to show^ first, that the two kinds of 

 variation must be ascertained ; and second, that the internal 

 organs must be reckoned with in an attempt to base the classi- 

 fication of the members of the genus on a firm footing. 



The question, then, which is raised by these considerations is. 

 What variation do the organs of Arenicola exhibit in differ- 

 ent specimens and in the growth of the individual ? When 

 the range of these variations is known, the systematic arrange- 

 ment of the growth-stages and the adults will be more soundly 

 based than heretofore. The gills and sette may first be 

 considered, as they are usually held to be diagnostic. With 

 reference to the gills we have shown that the common lug- 

 worm has two varieties of the gill, a pinnate and a dendritic 

 type. The pinnate type also occurs in A. cristata and less 

 conspicuously in A. Clap are dii. These three species cannot 

 then be separated by appealing to the form of the gill-plumes, 

 though as a group they are clearly separated by their common 

 type of gill from the two remaining forms, A. Grubii and 

 A. ecaudata, in which the mode of branching is quite dis- 

 tinct, but which agree so closely inter se in the matter as 

 to be indistinguishable on this character alone. With regard 

 to the development of the gills, different post-larvalindividuals 

 of A. marina vary in the time of origin and relative degree 

 of growth of these organs, but in all cases the gills appear to 

 arise in the centre of the future branchial region and spread 

 forwards and backwards. Accordingly until the full number 

 is attained it would be impossible to separate any of the three 

 caudate species of Arenicola by the position and number of 

 their gills. 



The setae show very much the same kind of variation. 

 Their form and characters are gradually changing through- 

 out life, so that there are very considerable differences 

 between the set« of a given species taken from specimens of 

 different age, diffei'ences more considerable in some cases than 

 between the sette of specimens of the same age but of different 



