NATURAL HISTORY OF SOME TUBE-FORMING ANNELIDS. 231 



in the midst is sent on down to the mouth by the converging currents from the 

 cilia. 



Amphitrite ornata has three pairs of gills, each gill having a main stem and 

 numerous dichotomous branches. The gills are almost constantly extending and 

 contracting, especially when the animal is lying in a tube. The evident purpose of 

 the movement is to bring in oxygen-bearing water. 



The nerve-action controlling the movement of the gills is not correlated; in fact 

 it is only occasionally that all the gill-stems contract or extend in unison. Although 

 the movement of the gills is not synchronous there is a regular periodic extension 

 and contraction of the large posterior gills, at least when the animal is in its tube. 

 The number of "respirations" per minute is fairly constant for each individual. Foi 

 example, one specimen repeatedly extended and contracted its posterior gills twenty 

 times a minute. Some specimens breathed as many as twenty-five to twenty-seven 

 times per minute. 



II. DIOPATRA CUPREA AUDOUIN ET EDWARDS. 



While digging in the sand and mud on Ram Island, Wood's Hole, Mass., in the 

 summer of 1901, I came upon a worm tube about ten inches in length. Six inches 

 of the tube was thickly covered with small pebbles and bits of shell. The remainder 

 seemed to be continuous with the lining of the roughly covered part, and was made 

 of thick, tough, leather-like material. On examining the tube at the laboratory 

 I found it to be inhabited. The annelid was a good specimen of Diopatra cuprea, 

 the natural history of which has been described at some length by Verrill, in his re- 

 port on "The Invertebrates of Vineyard Sound." Being curious to know how the 

 tube came to be covered with pebbles and shells while the animal lived in a region 

 of fine sand and mud, I placed it in a vessel of sea water, and distributed over the 

 bottom the large pieces of its old tube with some other material. In a very- 

 short time the annelid began to form a new tube. I watched the operation closely 

 during the formation of this tube, and found that the tube-forming activities agree 

 in outline with the description given by Verrill. As that author says, the animal 

 crawls partially out of its tube to collect pebbles, and then places them in position 

 at the margin of the forming tube. The stages in this process and certain variations 

 of method were not detailed by Verrill nor by any of the older observers, so far as 

 I know; hence I have thought it worth while to record them. 



The tubes of Diopatra in the region about Wood's Hole are frequently over twelve 

 inches in length. From one-third to two-thirds of this length is composed of a sand- 



