REPORT ON THE RADIOL ARIA. 1091 



The fjimily A n d r o s p y r i d a differs from the three preceding families of S p y r- 

 o i d e a in the three-jointed shell, in which the galea of the Tholospyrida and the thorax of 

 the Phormospyrida are combined ; it may therefore be derived from the latter by develop- 

 ment of a galea, or from the former by formation of a thorax. The common ancestral 

 group of these three famihes is probably the Zygospyrida. But it may be that a 

 part of the Androspyrida has originated directly from the S t e p h o i d e a. 



Two species only of this family were hitherto known, Amjihispyris reticulata 

 (figured, in 1872, byEhrenberg, as Dictyospyiis) and Sphcerospyris sj^/iogra (described, in 

 1882, by Biitschli, as Dictyospyris). Thirty new species were found in the collection 

 of the Challenger, and are here dis^^osed in seven genera. But perhaps these may 

 represent three diflferent families, which we here enumerate as subfamilies ; the Lampros- 

 pyrida, Perispyrida, and Nephrospyrida. 



The Lamprospyrida, the first subfamily, are tripodal Spyroidea with galea and 

 thorax ; they may therefore be regarded as a combination of Tholospyris and Acrospyris. 

 In the simpler form, Androspyris (PL 83, fig. 8), the network of the shell is simple, and 

 allows a very clear recognition of the internal stritcture. In the splendid Lamp>rosp)yris, 

 derived from it (PL 89, figs. 13, 14) the delicate network is more or less spongy. The 

 large apical horn is branched and fenestrated in the latter and simple in the former. 



The Perispyrida, the second subfamily, possess no free apophyses, but a three-jointed 

 shell wdth two distinct transverse strictures, an upper mitral and a lower coUar stricture. 

 The network of this shell is incomplete in Amphisjyyris (PL 88, figs. 2-7), being- 

 developed only in the frontal or lateral perimeter, thus making it prolial^le that this 

 subfamily may have arisen directly from the Tympanida {Toxarium, PL 88, fig. 1 ; 

 PL 93, figs. 18-20). The network of Tricolospyris (PL 88, figs. 8-11) is complete, of 

 another shape in the cephalis than in the galea and thorax; it is doubled or spongy in 

 Perispyris (PL 88, figs. 12, 13). 



The Nephrospyrida, the third subfamily, contains two very different genera only, 

 each of which may represent a separate family. Sphwrospyris (PL 83, fig. 4) possesses 

 a spherical or sulxspherical lattice-shell, the central part of which includes a sagittal ring, 

 connected with the inner face of the sphere by a number of paired apophyses, correspond- 

 ing to those of Lithocircus ; it may be easily confounded with some Sphseroidea 

 {Caiposphcera), but the enclosed ring leaves no doubt as to its true origin. Nephrospyns 

 (PL 90) has a flat discoidal shell of considerable size, usually kidney-shaped or cordate, 

 with a deep sagittal incision at the base, indicating the region where the pseudojDodia 

 are protruded from the porochora. The violin-shaped central capsule has two large 

 lateral lobes on each side of the including sagittal ring, and the transverse nucleus is 

 perpendicular to its sagittal plane (PL 90, figs. 7, 10). Sphcerocircus Si& w&\\ asNephro- 

 spyris have probably arisen independently from the Stephanida. 



