LOWEK CAMBRIAN. 51 



side in the casts, which probably is due not to differences in the animal itself, but to 

 differences in the manner in which the casts were produced. 



Another form (figs. 11-14) has more the appearance of a sea star, but is perhaps 

 related to the preceding. It has mostly four, but sometimes five arms, which vary in 

 form and size, but ordinarily agree in this, that one side is nearly flat, while the other 

 is strongly convex or more or less distinctly keeled. This form is found at Lugnas under 

 the same circumstances as the preceding. For the present it would be improper to 

 designate it by a special name, because the limits of the species can not be defined. 



These two forms are the most prominent among the echinoderms of the Eophyton 

 sandstone, and the others may possibly be referable to them as accidental variations, 

 due to the mode of fossilization. A detailed description of the latter without figures 

 would afford but little information, and may, therefore, be for the present postponed. 

 The feature which besides their age renders these fossils most remarkable is the 

 alternation between four- and five-parting. 



Dr. Nathorst's study resulted in an entirely different interpretation of 

 the zoological relations of the species. His description is as follows: 



A species belonging to the acraspedote medusae, in which either four- or five 

 partition prevails. Gastric cavity pyramidal, with four sided, roundly five-sided or 

 round base, sharply defined grooves on the sides; from the middle of the lower side 

 of the umbrella there projects downward into the gastric cavity a small round spike. 

 The genital apertures between the arms communicating with the gastric cavity. 

 Mouth opening pyramidal, four- or five-edged; arms, at least in young specimens, with 

 longitudinal open grooves, not branched (?). Presumably long tentacles. * 



Of course I had great hesitation in referring all these forms to only one species, 

 and I must insist expressly that I do not mean thereby to say that they do not com- 

 prise several. The latter seems even more probable, but the agreement in the organi- 

 zation of the parts that are here preserved, though possibly belonging to widely differ- 

 ing medusa?, does not allow this question to be settled with certainty. There might 

 be a better chance for arriving at a decision if we had definite data regarding the 

 occurrence of the various specimens in the rocks, and on the respective upper or lower 

 sides of the strata. At present, on the contrary, more regard must be had to the 

 difficulty of drawing definite boundary lines between the various forms, and so forth. 



How great is the difficulty really met with in the attempt to define the number 

 of species may best be understood if we imagine some tons of medusae of various kinds, 

 but not rhizostomids, cast up on a beach and leaving impressions of their mouth open- 

 ings. * * * Now, as it seems not quite impossible to refer the various forms to one 

 common type, I judged best, at least for the present, not to separate them, so much 

 the more as various differences may be due solely to phenomena of concretion and 

 other causes. From a theoretic standpoint, it might seem most reasonable to assume 

 the contrary, for seeing that a perfectly typical acraspedote medusa is found, there is 

 good reason for expecting more. The main point, of course, is that the presence 

 of acraspedote medusae in Cambrian time can in any case be affirmed with perfect 



