382 HERBERT P. JOHNSON. 



as in Lycastis quadraticeps, they have no discoverable relation to 

 each other; in other forms [Ophriotrocha pnerilis, Hesione sicula, 

 Sabella microphthalmia), there is a definite hermaphrodite organ 

 (Zwitterdruese). In the other and more exceptional condition 

 the individuality of the somite asserts itself by the production of 

 either male or female gonads. This is best seen in some of the 

 serpulids (notably Spirorbis), ova alone being produced in a few 

 of the most anterior abdominal somites, and sperm-cells in a 

 larger number of posterior ones. It is obvious that a combina- 

 tion of this condition with fissiparity, if the plane of division leaves 

 all the male gonads in one zooid and all the female gonads in the 

 other, leads directly to an alternation of .generations in which an 

 hermaphroditic parent resolves itself into unisexual offspring. 

 According to Haswell ('86) this actually occurs in Syllis corrus- 

 cans, in which the anterior portion (ordinarily forming the asexual 

 stolon in Syllids) produces ova, and the posterior somites separate 

 as a male worm. 



The ova are almost always small or even microscopic among 

 the Polychaeta. The egg oi PI aty nereis dumerili, measuring .41 

 mm., has been regarded as one of' the largest. Recently, how- 

 ever, Wiren ('07) has found those of Macellicephala to be much 

 larger, and perhaps of maximum size in the entire subclass (.76 

 by .48 mm.). While the ova of Lycastis quadraticeps are not so 

 large absolutely, measuring .43 by .28 mm., relatively to the 

 size of the parent they are far larger. 



There are at least two other species of Polychaeta which pro- 

 duce very few relatively large ova, — Nerilla antennata Schmidt, 

 '48 [Dujardinia rotifera of de Quatrefages, '65), and Amphicorina 

 cursoria de Quatrefages, '65. Both of these species produce 

 ova not only relatively and absolutely of unusual size, but ex- 

 tremely few in number (de Quatrefages's figures, give six in 

 Amphicorina and only four in Nerilla), and apparently they con- 

 stitute but a single brood. In these two species therefore, we 

 find the same condition of things as in Lycastis quadraticeps — 

 reduction in the number of eggs concurrently with increase in 

 their size — and the process has gone still further than in Lycastis. 



That this change is not coordinate with complexity of organiza- 

 tion or any real advance towards a higher plane of being, but 



