24 FOSSIL MEDUSA. 



2. The influence of the condition of preservation is readily seen in the 

 form of the umbrella in distorted and broken specimens, but it is not so 

 easily determined with relation to the presence or absence of the gelatinous 

 matter that is so prominent in the umbrella of recent species of the Acras- 

 peda. In some indivi duals it appears to be present (PI. I, figs. 1, 2, 3, 4), 

 while in others the firm, supporting ectoderm appears to have shrunk so as 

 to leave only a skeleton form of the umbrella (PI. I, fig. 7 ; PI. II, figs. 5 

 and 6). Frequently the medusa is completely embedded in the siliceous 

 matter of the nodule, and it is only by cutting sections that its form can be 

 observed. The embedding may be entire (PI. IV, figs. 10, 11, 13) or partial 

 (PI. Ill, figs. 2, 9; PI. IV, figs. 1, 3, 5, 8). 



The lobation of the umbrella is rarely, if ever, lost; it is the dominant 

 chai'acter in all the specimens, and extends from or near the center to the 

 margin and forms the lobate marginal border (PI. I, figs. 1, 2, 4), which is 

 often deeply indented (PI. Ill, figs. 1, 5, 6). The lobes vary in number 

 from 6 to 20, or more, and in form from broad, slightly rounded to narrow 

 and strongly rounded. There is no regular sequence of 6, 8, 12, etc.; on 

 the contrary, the irregular numbers 5 and 7 are largely represented (PI. I, 

 figs. 4 and 5 ; PI. Ill, fig. 1), and 6 and 8 are abundant. The sinus between 

 the lobes may be merely a depressed line (PI. I, figs. 1, 2, 3), or it may cut 

 through to the base, leaving only the central portion of the lobes attached 

 to the umbrella disk (PI. 1, figs. 5, 7, 8; PI. Ill, figs. 3, 5, and 6), or it may 

 be irregular, varying in size and arrangement (PL I, figs. 7 and 8 ; PI. II, 

 fig. 5). In the former instances the umbrella is symmetrical and relatively 

 smooth, while in the latter the surface is broken by the normal series and a 

 secondary series of lobes that may be represented in an individual by a 

 single narrow lobe projecting from between and beneath two of the upper 

 lobes (PI. I, fig. 6, x), or by a much larger number, as seen in PI. II, fig. 1, 

 where eight project from beneath the upper series. The variation in the 

 latter group is considerable and is not reducible to any serial tabulation. 



The more regularly lobed individuals may be taken as the first type 

 of the species (PI. I, figs. 1, 2, 3, 4), and the irregularly lobed as a variety 

 (PI. I, figs. 6 and 7; PI. II, figs. 1 and 3). So many gradations occur 

 between the two extremes (fig. 4 of PI. I, and fig. 1 of PI. II), that I do not 

 find it practicable to distinguish two species or even varieties. The passage 

 from the typical forms (PI. I, figs. 1 to 4) to the variety represented by figs. 



