200 CARNEGIE INSTITUTION OF WASHINGTON. 



loose rocks on the surface of the reef. The only one of the latter that was at 

 all abundant was a slender, delicate species Hving in tubes, the tubes being 

 composed of small stones loosely cemented together. An extremely common 

 species was first regarded as a Nicidion, but a study of a more complete 

 series showed that they are evidently very young individuals of Leodice 

 viridis, the palolo. There is no record of the palolo swarming at Suva, and 

 I got no well-developed individuals. From a comparison of these with speci- 

 mens collected in Samoa, I am convinced that the identification is correct, 

 and that, if it swarms at all in Suva, the numbers of individuals are too small 

 for the swarming to attract attention. It seems to me possible that, because 

 of unfavorable conditions, the animal may never reach sexual maturity on 

 the inner parts of the reef, the population being kept up by migration in the 

 trochophore stage from other localities. 



In American Samoa a thorough study was made of the annelid fauna of 

 the reefs in Pago Pago Harbor. Here, as in Suva, the coral rock is rather 

 poor in annelida, but the Leodicidse are represented by the greatest number 

 of species. In this harbor, as in other parts of the islands, the mud-flats lie 

 at the mouths of streams, where the dilution by fresh water prevents the 

 growth of Polj'chsetes, and a capitellid is usually the only genus represented, 

 though some terebellids were collected at the upper end of Pago Pago Harbor. 

 A Lumbrinereis and an Arabella were collected on the under side of stones 

 from Utelei and Aua reef respectively. This is of interest, for these are charac- 

 teristically mud-living forms, and their adaptation to this peculiar habitat is 

 very unusual. Species of Leodice, Lysidice, Nicidion, Stauronereis, and 

 Aglaurides were collected in the coral rocks, the first-named genus being the 

 most abundant. Many of the species are identical with those collected in 

 Suva, L. viridis, the palolo, being the commonest. These are found in all of 

 the reefs, being more abundant in those regions where the annual swarming 

 is most in evidence, though I found no locality where they were abundant 

 enough to give rise to the enormous numbers found in the annual swarm. 

 Since the larger individuals appear toward the outer margins of the reefs, 

 I am of the opinion that the greater number of the animals live in the deeper 

 parts of the reef, on the outer margin, where, on account of the surf, col- 

 lecting is practically impossible. Toward the shore, the individuals become 

 very small and undeveloped, though I could discover no other difference 

 between these and the large ones. If our information about the swarming is 

 correct, the eggs are all laid at the same time, and the individuals found in 

 1920 were hatched from eggs laid not later than November 1919. As stated 

 above, of the Suva specimens, I question whether those nearer shore ever 

 become sexually mature, because of the unfavorable environmental conditions. 



A comparison in L. viridis was made between those collected at the spring 

 tide in May and the corresponding tide of June. The latter were distinctly 

 farther advanced toward sex maturity than the former, though in only a few 

 cases were the eggs sufficiently mature to begin to show the green coloration. 

 It should be remembered that the latest collecting was several months before 

 the swarming, so that no very high degree of maturity was to be expected. 

 Dr. Mayor, director of the expedition, kindly collected for me during the July 

 spring tide, but found there Avas very little change in maturity over the con- 

 dition occurring in June. 



The reef at Aunuu village was fairly rich in annelids, the most abundant 

 being a large species apparently identical with one collected at Suva. That 

 at Leone was very poor, and the mud-flats inside the reefs are too abundantly 

 supplied with fresh water to be favorable to annelid growth. The most 

 characteristic mud-flat on the island is along the eastern shore of the lagoon 

 near Nuuli, where a small Marphysa was found. 



