THE SPONGES OF THE WEST-CENTRAL PACIFIC 153 



between 3 and 13 mm are more common. Lateral growth occurs indefinitely. 

 For example, specimen No. M. 141 completely covered a large, dead, ramose 

 coral. 



The exterior color in life was always a vivid red. This bright color 

 extends to a depth of 1 to 3 mm in the sponge. Whenever specimens were 

 more than 3 mm thick, the deeper portions were more orange than red. The 

 consistency was primarily softly colloidal ; but, inasmuch as the spicules could 

 be felt by the fingertips, the adjective "gritty" was often applicable. 



The surface is regularly smooth and usually lipostomous. Only under 

 exceptional circumstances were specimens brought so quickly to the surface 

 and into preservative that the pores were not first closed. Specimen No. M. 

 365 shows them to be 100 /x to 200 /x in diameter, about 5 per each square mm. 

 Even in this specimen, the oscules cannot be distinguished from the inhalant 

 openings. 



The ectosome is characterized by a strong development of protoplasmic 

 structures not sharply separated by extensive subdermal spaces from the 

 underlying endosome. The latter is permeated by a fibro-reticulation, but 

 this is very openwork with large areas of fleshy material in between the 

 scattered tracts. 



The skeleton consists, as noted above, of fibers which range from 90 /.<. 

 to 150 /x in diameter and contain some spongin and many spicules. The 

 megascleres, which are tylostyles, are astonishingly uniform in all the speci- 

 mens from the many regions noted, ranging from 10 /x by 520 /x to 13 /x by 

 490 /x in size. There are typically seven microsclere sorts. First there are 

 raphides, almost invariably about 40 xi long, although in Specimen No. M. 

 234 they were only 12 xi long. There are regularly two sizes of sigma, the 

 larger 45 ti to 50 ti in chord length and smaller 15 /x to 22 /x in chord length. 

 These are also very similar in all the specimens. There are four types of 

 chela, always palmate anisochelas. The largest size is invariably present, and 

 in no case were they found (in the summer of 1949) to be arranged in 

 rosettes. Very often they were found in the echinating fibers, the small 

 end embedded and the large end protruding from the fiber. This spicule 

 category ranged from 122 x<, to 140 xi in total length with very little variation 

 between specimens. The next largest size was 70 /x in length ; it was missing 

 from about half the specimens studied. The third size was 25 xi in length. 

 It was missing from about half the specimens, although not the same half 

 in every case. The fourth and smallest size of chela varies from 12.5 /x to 

 16 /x in length and is almost always present. No specimens were found with 

 less than two of these four kinds of chela. When so few occurred, one was 

 always the largest size, but the second might be any of the other three kinds. 



This species was first described as Mycale armata by Thiele 1903, page 

 950, from the East Indies. It is highly probable that several species since 

 described should be reduced in synonymy to armata, but published descrip- 



