124 THE BERMUDA ISLANDS. 



is different from that wliicli Duchassaing and Michelotti ob- 

 tained at St. Thomas and Guadeloupe, relying probably on 

 the discrepancies between tiie poor figures given by the latter 

 authors and the more correct one wiiich Gray has given. He 

 consequently retains the specific term tuherculatus for Gray's 

 form; proposing for Duchassaing and Michelotti's the name 

 Duchassaingi. There is little room for doubt, however, that 

 the two forms are identical; my observations have shown 

 that the species is to be referred to the genus Mammillifera as 

 defined by Erdmann. 



The specimens from the Bermudas were either solitary, 

 attached to a piece of rock by a base only very slightly ex- 

 panded, or else were grouped together in twos or threes, in 

 which case they were united by a slightly-developed, flat or 

 slightly tubular coenenchyma. In none had the ccEnenchyma 

 any such tubular or stolon-like form as is shown in the figure 

 given by Duchassaing and Michelotti. Judging from the speci- 

 mens I studied, the tendency to form a coenenchyma is slight. 



The polyps (PI. 11, fig. 5) vary in height from 1.3-2.7 cm. ; 

 their diameter being about 0.7-0.9 cm. The column is marked 

 by six or eight di.stinct annular grooves, and by from twenty 

 to twenty-five longitudinal ones. In the lower part of the 

 column the ridges formed by these longitudinal grooves are 

 entire, but higher up they begin to be divided into series of 

 tubercles, a row of these corresponding to each ridge. These 

 tubercles increase in size towards the margin and several 

 become grouped together upon elevations of the column wall, 

 giving rise to mulberry-like protuberances. Near the margin 

 the tubercles suddenly cease, forming, in contracted specimens, 

 a strong ridge bounding the dome-shaped area which forms in 

 such specimens the summit. This dome-shaped area belongs 

 to the column, the animal being strongly contracted, and 

 though without tubercles shows clearly the continuation up- 

 wards upon it of the longitudinal furrows, and is, accordingly, 

 marked by a series of radiating ridges. 



In structure the tubercles of the column are solid, being 



