THE SPONGES OF THE WEST-CENTRAL PACIFIC 265 



anthozoan skeleton ranging in size all the way from minute bits up to slabs as 

 large as table tops. The crest of the reef also consists quite extensively of the 

 same sort of dead material — sometimes loose pieces akin to gravel, or shingle, 

 or fragments small enough to be called sand. Again a cementing process has 

 made a sort of coquina or calcareous sandstone. 



Three sorts of openings exist for transfer of water into and out of the 

 lagoon. First, there may be one or a very few wide, deep openings, perhaps 

 100 meters wide and 30 deep, through which the tidal currents run at 2 to 

 (rarely) 4 kilometers per hour. It will be noted that such channels are richly 

 lined with live coral and other sessile life. Second, there may be one or a few 

 similar but smaller channels, say 20 meters wide and 10 meters deep. The 

 tidal currents in these are often more than 5 kilometers per hour. The sides 

 and bottoms of them are scoured fairly clean of life. Third, there are usually 

 dozens of smaller channels over the reef, often so shallow that they are dry 

 at low tide. At half tide, the currents in these are quite violent. They, too, are 

 devoid of all conspicuous forms of sessile life. 



On the Marshallese atolls, near the southwest side of the lagoon (the 

 prevailing winds being from the northeast), there is a small, often somewhat 

 incomplete reef 100 to 300 meters off shore, enclosing a miniature lagoon 

 within the greater lagoon. 



Among the principal algae which construct the reefs are Halimeda and 

 Lithothamnion. In places algae actually do more to build up the reef than do 

 coelenterate corals. My own observation, however, was to the effect that 

 Anthozoa were more conspicuous participants in the life of the reefs. In some 

 areas "staghorn" corals, probably of the genus Acropora, form dense thickets. 



A comprehensive report on the madrepore and millepore corals of the 

 Western Pacific may soon be published by Dr. John W. Wells. Probably 

 such reef-building genera as Stylophora, Orbicella, Pontes, and Pocillopora 

 may be reported. 



It will be necessary to speak of sponge abundance in generalized terms, 

 because methods which might yield precise returns for other animals are 

 inappropriate for the Porifera. 



Considering the numbers of individuals is valuable as a factor but mean- 

 ingless alone. There are areas with hundreds of specimens of Spirastrella, 

 but each just a fingernail-sized fleck, and such an area is best described as 

 poor in sponges. 



The same applies for the size of individuals. In a whole bay there may 

 be many kilograms of sponge, but all aggregated in the form of one or two 

 huge specimens of Stellettinopsis. This is not abundance of Porifera. And how 

 shall one measure sponge quantity? A freshly removed specimen is mostly 

 water, which does not readily drain out of the complex system of canals and 

 chambers. The bulk of the dry weight of many sponges is dead skeleton — 

 silica or calcium carbonate or spongin. 



