459 



sensory structures but to aid in drawing food into the mouth. In the 

 present case however only the proboscis was pushed into the anemone 

 and under such circumstances the cirri can be of no service whatever 

 in the collection of food. In one specimen oi Rhopalome?iia in my pos- 

 session the cirri project some distance through the mouth opening and 

 rather support the view that these organs are sensory entirely, enabling 

 the animal to determine the presence of food or the nature of its sur- 

 roundings. 



During the past summer while acting as naturalist on board the 

 Stmr. Albatross in the waters of Alaska, I had the opportunity of ob- 

 serving the habits of two species of solenogastres. The first belonging 

 to the genus Chaetoderma was comparatively abundant in the tenacious 

 glacial mud at a depth of about 300 fathoms and several specimens 

 lived for several days in captivity. A tall jar was nearly filled with 

 mud and immersed in a larger aquarium into which a stream of water 

 was continually running. The specimens were the placed on the sur- 

 face of the ooze and within a short time they commenced to burrow. 

 This was accomplished almost entirely by movements of the prothorax 

 essentially like those of the front end of an earthworm when in the 

 act of burrowing. The operation was initiated by decreasing the cal- 

 ibre of the prothòrax as much as possible whereupon it was pushed 

 into the mud and then expanded to twice the minimum diameter. 

 Judging from two specimens carefully observed the process occupied 

 about 15 seconds and was at once repeated. With some other speci- 

 mens it was much slower due perhaps to rough handling in the dredge. 

 One individual, about 41 mm in length, disappeared from view in one 

 hour and twelve minutes; another of about the same size vanished 

 thirty -two minutes later; while others required all the way from three 

 to nine hours. 



In these movements the anterior third of the prothorax is the 

 most active, often swelling to twice the greatest diameter of the remain- 

 der of the prothorax. At this time its trumpet-shaped appearance al- 

 most exactly duplicates the familiar drawing of Chaetoderma nitidulum 

 by von Graff. While the extreme front end of the animal is thus 

 most active in the excavation process the remaining parts of the pro- 

 thorax also take a part. The same is true of the body proper. It is not 

 passivly dragged into the opening but is forced in by slow vermicular 

 or peristaltic movements that travel the entire length of the animal. 



Some of the specimens, generally the weaker, excavated relatively 

 short tunnels from which they extended the posterior end of the body 

 with the gills expanded. The more active individuals tunneled much 

 more extensively; in fact some of them appeared to be at work contin- 



30* 



