ARISTOTLE'S LANTERN AND PERIGNATHIC GIRDLE. 189 



In the Centrechinoida, as an order, there are pits in the top of the pyramids beneath the 

 epiphyses, the foramen magnum is always deep, styloid processes are visible in face view, and 

 the lantern is erect excepting in Astropyga and the Echinothuriidae, where it is inclined as in 

 the young, and the adult of Palaeozoic types. Other characters of the lantern in this order 

 are sufficiently distinctive to justify the formation of three suborders, as shown (p. 183). In 

 all the above the lantern is made up of forty pieces, braces and compasses are nearly alike, 

 and corrugations exist on the pyramidal wings for attachment of interpyramidal muscles. 



In the Holectypina the structure of the lantern in Discoidea was shown in part by Loven 

 (1892), and has recently been shown very completely by Hawkins (1909) in Discoidea cylin- 

 drica. As Hawkins shows, the teeth are narrow, elongate, and keeled; pyramids are wide- 

 angled with a rather deep foramen magnum and laterally with corrugations for the attachment 

 of interpyramidal muscles ; the epiphyses are relatively large but narrow and closely resemble 

 those of the Stirodonta. In the whole structure the lantern approaches much nearer to that of 

 the Stirodonta than of any other group of Echini. Of course it may differ in other genera of 

 the Holectypina where it is at present unknown (see p. 218). 



In the Clypeastrina the lantern is procumbent, but is more erect in the young, as I have 

 seen in Echinarachnius parma. The teeth are keeled, the foramen magnum is extremely 

 shallow (text-figs. 214-216), pyramids are highly modified with wing-like extensions without 

 ridges for the attachment of interpyramidal muscles, or pits in the top of the pyramids; brace 

 and epiphyses are very small, and compasses are absent. The whole lantern is much 

 specialized, the keeled teeth presenting the best connecting character (see p. 218). 



The Spatangina as a suborder are characterized by the absence of the lantern, and it is 

 of highest interest that a lantern exists in young Echinoneus, as shown by Mr. Agassiz (1909). 

 In a very early stage this type has a lantern with keeled teeth, deep, open foramen magnum 

 and bifid compasses, the retractor muscles are attached to the ambulacral auricles, and the 

 protractor and radial compass muscles are attached to the base of the single primordial inter- 

 ambulacral plates. It is interesting and suggestive that this combination of characters shown 

 by Mr. Agassiz in young Echinoneus finds an exact parallel in adult Arbacia (text-fig. 212, 

 p. 184; text-fig. 227, p. 193), and further, the Stirodonta, as here defined, is the only group 

 of regular Echini that does possess such a combination of characters (p. 217). 



Before considering the perignathic girdle and the relation of the lantern muscles to the 

 same, I would call attention to the relation of the lantern <ind muscles to the test as a whole. 

 In the Centrechinoida, as Loven (1892, p. 44) said, "the retractors of every pyramid are at- 

 tached on different ambulacra, each on a rising auricular branch." He considered the muscles 

 arising from a single pyramid, the halves of which are united by suture, whereas I take the two 

 half-pyramids united by the interpyramidal muscles as the basis of consideration. The dis- 

 tinction is purely a morphological one. In other words, a pyramid may be considered as inter- 



