598 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



The position of the ventral surface appears to be a matter of complete in- 

 difference; it is rarely horizontal, and if so, is as often directed downward as 

 upward. 



On the reefs it has been observed that the crinoids usually occupy restricted 

 localities and are not by any means universally present. Thus at Maer Island, 

 Torres Strait, Dr. H. L. Clark found them on different parts of the reef flat; but 

 except for the ubiquitous Comatula fwjmrea they were virtually confined to the 

 southern and eastern side of the island. They were usually in water not less than 

 2 or 3 feet deep at low tide, but occasionally when the tide was very low they were 

 in water only a few inches deep. 



Comatulids have been reported, in very varying abundance, from a large num- 

 ber of reefs all over the Indo-Pacific region, from southeastern Africa, Mada- 

 gascar, and Mauritius to the Red Sea, and eastward to northern Australia and 

 Oceania; but they occur in the greatest variety and abimdance on the reefs of 

 northern Australia and the Malayan region, including western Oceania. 



On the reefs in the Atlantic they are very rare, both on the eastern and western 

 shores, having been recorded from only about a dozen widely separated localities 

 (including St. Helena), in all cases but two (St. Helena and Eio de Janeiro) only 

 once, and in all cases but one only a single species was reported. 



In water of a few fathoms' depth along the outer side of the reefs comatulids 

 often occur in very great variety and abundance, and it is in such situations that 

 the large multibrachiate species are found. Thus the littoral reef faima is, prop- 

 erly speaking, the upper fringe of a much more extensive fauna existing from the 

 shore line down to between 20 and 30 fathoms. 



It is understood that the term reef as used above is employed in a somewhat 

 broad sense. In most cases it refers to the " coral " reefs properly speaking, but in a 

 few cases the reefs are of volcanic rock, with a superficial and sparse and often 

 restricted growth of coral. 



In many places in the Mediterranean and Adriatic and in a few localities on 

 the Atlantic coast of Europe Antedon occurs on partially submerged ledges and 

 along rocky shores which are more or less furnished with a heterogeneous growth 

 of marine organisms. 



Occurrence of comatulids among seaweeds. — In certain places on the Atlantic 

 coasts of Europe Antedon bifida occurs in considerable numbers among the roots 

 of rockweed. 



Dr. L. E. Griffin reports an unidentified species from Culion, Philippine 

 Islands, as being a very active swimmer and as living among the eelgrass. 



At Tobago Dr. H. L. Clark found that the bottom frequented by Tropioinetra 

 picta is made up largely of Porites fragments, usually more or less covered by a 

 growth of Gorallina and Hcdim^da; scattered over it there is also a sparse growth 

 of short eelgrass (Zostera). Sometimes the bodies of the animals are more or less 

 completely shaded by a clump of eelgrass or seaweed, but this is not usually the 

 case. Generally the individuals are solitary, but occasionally five or six may be 

 foimd about a single cliunp of Porites or of eelgrass. 



