A STUDY IN VARIATION. 51 



and slender. Neither the right, nor the left antenna is permanently 

 responsible for the furnishing of the opercular stopper, for out of 

 92 specimens examined, in 50 cases the right antenna i.e. that one 

 belonging to the right branchial fan performed the opercular duty, 

 having the stalk elongated and the extremity swollen into the usual 

 opercular plug. In the remainder, viz. in 42, the left antenna was 

 become the functional operculum. 



This fact that in practically equal number of cases, the operculum 

 is developed indifferently from either the right or the left antenna, 

 is obviously of extreme importance in any attempt to trace the story 

 of its origin and subsequent history. Before making this attempt, , 

 we will proceed to notice the more important of the aberrant forms of 

 antenna and contrast them with the normal form, and with its 

 homologue, the functional operculum. 



The Operculum : The normal form is doubly infundibuliform, 

 fairly well figured by Johnston (1) for his Serpula reversa, which 

 I am convinced is the same species as the present.* The upper cup 

 is beset with 12 to 17 bipinnate spinous processes, while the lower 

 has a multiserrulate margin, the serrations being continued as grooves 

 for some distance downwards on the inside as well as on the outside 

 (PI. v, Fig. 11). 



The form is subject to not inconsiderable variation within certain 

 limits. Thus the larger specimens count more pinnate process than 

 the smaller, and the relative proportions of the two cups are 

 inconstant. 



One case was however very remarkable, for the lower cup was 

 absent, and represented solely by a large elongated swelling as shown 

 in Fig. 14, while the processes which represented the pinnate spines 

 of the normal upper cup, were merely rather deep crenulations with 

 smooth margins. Fig. 12 and 13 show 7 intermediate stages in this 

 variation, showing in Fig. 12 a great swelling of the lower cup and 

 a decrease in size in the pinnate processes above, while in Fig. 13, 

 these spines are still further reduced and all trace of the crenulated 

 margin of the lower cup lost. In this case also, the part representing 

 the lower cup is much swollen. Figs. 12, 13, and 14 are therefore 

 perfect gradational steps in the retrogression in shape of a normal 

 operculum (Fig. 11). 



The Antenna : The normal form is i to | the length of the 

 operculum. The average length of those of adult specimens of 



* The only differences I can trace are that S. reversa is solitary and has more 

 branchiae and more lateral teeth on the spinous processes of the opercular cup ; 

 S. pectinata occurring in colonies and being on the average smaller than S. reversa. 

 But I have traced all intermediate grades, so that the differences are not great 

 enough to constitute more than a variety. Fig. 15 shows typical form of operculum 

 of S. reversa. 



