DEPARTMENT OF MARINE BIOLOGY. 



221 



by any disintegrating agency, annelids occur. The large loggerhead sponges 

 contain many Syllidse, and one leodicid (Leodice conglomerans) occurs only 

 in these sponges, living inside a parchment tube of its own construction, 

 which branches to follow the cloacal system of the sponge. In a few species 

 the cavity occupied by the annelid passes through the solid rock in such a 

 fashion as to indicate that it must have been excavated by the animal itself. 



Certain of the serpulids and sabellids occupy tubes which open at the 

 surface of the living coral rock, but in this case it is quite evident that the 

 tube of the serpulid has begun its groA\i:h at the same time as that of the coral, 

 and while the growth has been about equally rapid in the two animals, the 

 tube has gradually been inclosed by the coral. Floating seaweed rarely 

 carries any annelids except, occasionally, a few small nereids, and Hermodice 

 caruncidata is occasionally brought to the islands on floating logs. Probably 

 the Leodicidge is the largest in number of individuals of any annelid family 

 represented here, with the Sabellidse and Serpulidse next. Nereidse are com- 

 paratively rare, occurring in the coral rock in about equal abundance with 

 representatives of the Capitellidae, Cirratulidse, Terebellidae, and Syllidse. 



Unless there may be more than one breeding season in the year, and evidence 

 on this point is as yet lacking, most of the annelids breed rather late in the 

 season. Leodice fucata, the Atlantic "palolo," swarms in coincidence with 

 the last quarter of the June-July moon, but there is no evidence of a similar 

 habit in any other leodicid, and specimens of most of the species I have studied 

 have been found still carrying large quantities of unripe eggs late in July. 

 Specimens of Onuphis (Paranorthia?) brought from the Marquesas spawned 

 in a live-car between July 12 and 15. In 1909 the only ripe annelid eggs I 

 could find prior to July 15 were those of Pomatostegus stellatus and Spira- 

 branchus tricornis. A small heteronereid was found swarming on July 12, 

 and another Nereis, rather common at Fort Jefferson, was found with unripe 

 eggs on July 24. Apparently the greater number of annelids in this locality 

 breed later than August 1. 



The swarming of the Atlantic "palolo," Leodice fucata, occurred on the 

 morning of July 11, the last quarter of the moon falling on the 14th. ^ This 

 swarming of annelids in coincidence with certain phases of the moon, formerly 

 supposed to be limited to the Pacific "palolo," Leodice viridis, is now, owing 

 to the work of Mayer,^ Izuka,^ Hempelmann,^ and Lillie and Just,^ known to 



^The following table gives the dates upon which the Atlantic palolo has been observed to swarm 

 at Tortugas, and the dates of the quarters of the moon. The date of the principal swarm is shown 

 in heavy type, while the dates upon which only a few worms were observed swarming are shown in 

 ordinary type. — A. G. M. 



*Mayer, A. G. The annual breeding-swarm of the Atlantic palolo. Carnegie Inst. Wash. 



Pub. 102, 1908. 

 ^Izuka, A. Observations on the Japanese palolo. Journal College of Science, Tokio, vol. 



XVII, 1903. 

 ^Hempelmann, Fr. Zur Naturgeschichte von Nereis dumerilii. Zoologica, Bd. 25, Heft 62, 



135 pp, 4 Taf.. 14 figs. 1911. 

 *Lillie and Just. Biological Bulletin, vol. 24, No. 3, February 1913. 



