450 THE OPISTHOBRANOHIATE 



The same statistics show that the veligers are carried out to con- 

 siderable distances from the shore and that after falling- to tlie 

 bottom and undergoing their metamorphoses they gradually make 

 their way to the shore. This year I found the first individual 

 which had so migrated on August 18th. at Drake's Island ; and a 

 few days later, as we were doing some collecting on the shore 

 at Bovisand, Mr. Woodward found two more. During July and 

 August I could not find a single large or mature individual either 

 on the shore or with the dredge ; and this leads to the conclusion 

 that the disappearance of the old individuals after the spawning 

 has been accomplished is due, not to a re-migration into deeper 

 water or to habits of concealment, but to death. Goniodoris nodosa 

 is an annual, and dies when it has ceased to deposit its eggs in the 

 spring and early summer (cf. Woodward, Manual of Mollusca, 

 4th ed., p. 12). 



Young specimens differ from full-grown individuals in several 

 points of structure as well as in size. The pleuropodial frill is rela- 

 tively lai'ger, and is generally freely scalloped at the edge. I have 

 given a representation of the animal at this stage (^ inch in length) 

 on PL XXVII (fig. 4). The points formed by the scalloping are to be 

 compared homologically with the filaments of Idalia and allied 

 forms : they generally contain special aggregations of opaque white 

 gland-cells, comparable with those of the pleuropodial filaments of 

 Ancula, Triopa, &c. (cf. Herdman and Clubb, 3rd Rep., pp. 136 and 

 184; Friele and Hansen on G. Danielsseni, 1. c, p. 72), 



The specimens of the so-called Doris Barvicensis of Johnston 

 which were found by Allraan among the roots of Laminaria digitata 

 in Courtmasherry Harbour in August and September, 1838 (see 

 Thompson, 'Ann. Nat. Hist.,' vol. v, p. 87), and the Goniodoris emar- 

 ginata of Forbes dredged in twenty fathoms off the Isle of Man in 

 October, 1839 ('Ann. Nat. Hist,,' vol. v, p. 105), were undoubtedly 

 young specimens of Goniodoris nodosa migrating to the shore. 



Another difference of considerable morphological importance be- 

 tween the young and adult Goniodoris nodosa is to be found in the 

 condition of the posterior portion of the pleuropodial frill. In young 

 specimens the lateral portions of this structure are invariably dis- 

 continuous posteriorly, as represented in my figure : this condition 

 is persistent in Goniodoris castanea throughout life, as it is in the 

 closely allied genus Idalia. But as the animal grows the basal 

 portions of the posterior terminations of these folds become con- 

 nected together, and give rise to a continuous circular fold like that 

 of Archidoris, which differs from the latter, however, in being deeply 

 notched or emarginate postei'iorly. This has been hitherto regarded 

 as the final character assumed by the fold in the species, and ex- 



