ON TIE METHOD OE DISPERSION . 



AND 



FERTILIZATION OF OVA IN SOME SABELLIDS. 



BY JAMES HORNELL, 



])RINCIPAL among the tube building Sabellids that live in 

 the Aquarium tanks at our Jersey Station, are a considerable 

 number of the fine Branchiomma (Sabella) vesiculosa, so remark- 

 able for well-developed eyespots at the tips of the filaments. This 

 species, one of the commonest on these shores, builds a tube some 

 7 to 9 inches long in the Zostera banks, with the upper opening on a 

 level with the surface, differing thus very markedly from the free 

 or projecting tube-form of the allied Sabella pavonia. 



' Towards midsummer, dissection showed the genital products 

 well advanced towards maturity, so seeing that this Sabellid does 

 not depart from the usual characteristic of the Polychaetes in having 

 the sexes separate, I was induced to watch very carefully for the 

 time when fertilization should take place. It is worthy of preli- 

 minary note that under natural conditions, the tubes are seldom 

 closely set — usually they are from two to six inches apart. The 

 plumes too have a very short radius and never make a broad far- 

 reaching fan-circlet as in S. pavonia. When on July 5th the 

 sexual elements appeared, the sight was most curious and worthy 

 of the long and constant watch. First, a female Branchiomma 

 was noticed passing up into the circlet of filaments, from — I believe — 

 the median gutter that runs along the veptral side of the body, rather 

 numerous large white ova. These gradually accumulated for some 

 time in the cone-shaped hollow formed by the filaments ; then 

 rather suddenly the animal retracted its plumes with considerable 

 force, such indeed that the ova by the impetus imparted were 

 shot upwards and outwards to a distance of several inches. A 

 minute's rest or less, saw the animal again with fully expanded 

 plumes, passing up ova again from below in rapid succession. Then 

 when the filaments were once more clogged, a similar sudden jerk 

 dispersed them widely ; so the process went on for nearly three- 



