CH. VI MABINE FAUNA OF TALIS SE SHORES 133 



dead coral stones. The Sarcophytums are extremely in- 

 teresting things as far as their anatomy and zoological 

 relationships are concerned, but I must confess rather 

 unpleasant things to look at and to handle. They are 

 Alcyonarians allied to the alcyoniums, common enough on 

 Enghsh coasts under the name of ' dead men's fingers,' 

 but instead of the polypes being all alike as they are 

 in Alcyonium, there is an interesting dimorphism to be 

 observed amongst them. Some of the polypes attain their 

 full development and act as the food-catchers and food- 

 digesters for the colony. These are called the autozooids. 

 The majority of the polypes, however, are arrested in their 

 development, and neve possess tentacles. These are called 

 the siphonozoids, and by means of a groove on one side of 

 their stomachs (siphonoglyphe) armed -^vith long powerful 

 cilia they act as pumps to produce currents of water, which 

 circulate through the comphcated system of canals, by 

 which the mass is perforated (32). The Sarcophytums are 

 like huge toadstools in shape, in many cases as much as two 

 or three feet across the disc. They are of a sickly yellow or 

 yellowish-green colour, and very slimy to the touch. In 

 the course of my wanderings I must have slipped down into 

 the water several dozens of times, scratching myself, tearing 

 my clothes, and frequently stinging my hands severely at 

 each disaster. Clavularia was very often the cause of my 

 loss of equihbrium, but the chief culprit was undoubtedly 

 Sarcophytum, whose broad, smooth, and slippery surface 

 never offered a sm-e foothold for one moment. Before 

 leaving Sarcophytum, I must note my disappointment at 

 never finding any of the autozooids properly expanded. The 

 polypes were frequently to be seen only partially retracted, 

 but I never saw them with their tentacles aU spread out, as 

 they are so frequently seen in Tubipora and Clavularia. I 

 could frame no theory that would account for the fact that 



