122 BULLETIN 77, UNITED STATES NATIONAL MUSEUM. 



new species, is exactly the same as in ClathroiJora, where many 

 students would probably refer both of them. However, in the latter 

 genus the zooecial structure is exactly the same as in Ptilodictya, 

 from which ClatJiropora is distinguished only by its cribrose or clathrate 

 zoarium. Granting that this is a sufficient generic distinction, 

 new genera would be necessary for the clathrate species of Grapto- 

 dictya and of Stictoporella, but I am unwilling to believe at present 

 that such genera would be of any advantage. The zooecia of both 

 Graptodictya and Arthropora differ from those of Ptilodictya in having 

 low peristomes and in being separated from each other by solid 

 interspaces bearing one or more tortuous, fine, elevated lines. 



Although Graptodictya has been defined by Ulrich and compared 

 with related genera on several occasions, no illustrations of the inter- 

 nal structure of the type species have ever been published. On account 

 of the close relationship between Graptodictya perelegans, the geno- 

 type, and the new species G. honnemai, I am taking this opportunity 

 to present the accompanying views of the former (fig. 47) to illustrate 

 the generic structure, as well as for comparison with the latter species. 



Genotype. — Graptodictya perelegans (Ulrich). Earliest SUurian 

 (Richmond) of the Ohio Valley. 



GRAPTODICTYA BONNEMAI, new species. 



Plate 8, fig. 3; text fig. 48. 



Cfr. Graptodictya (Ptilodictya) perelegans (Ulrich), Joiirn. Cincinnati Soc. Nat. 

 Hist., vol. 1, 1878, p. 94, pi. 4, figs. 16, 16a. 



The close relationship existing between this new species and the 

 genotype has been noted under the generic remarks. The internal 

 structure of the two is practically identical, and, were it not for cer- 

 tain differences in method of growth, they could hardly be distin- 

 guished specifically. As it is, Graptodictya honnemai is most certainly 

 the forerunner of G. perelegans, and the occurrence of such similar 

 species in the Middle Ordovician and earliest Silurian strata is but 

 another instance of the close relationship between the faunas of these 

 two periods. 



Critically compared, the zoarium of G. honnemai is found to branch 

 much less frequently than its later representative. Each species 

 branches dichotomously at regular intervals; in G. perelegans this 

 branching occurs at intervals averaging 4 mm. ; in the present form 

 10 mm. is the usual distance between the bifurcations. This gives a 

 more dehcate aspect to the Russian form. Other differences may be 

 noticed in the shape of the apertures, which in G. honnemai are, as a 

 rule, more elongate, but this is a character which varies with age 

 and must therefore be used with caution. 



The dichotomously branching, slender, ribbon-Hke zoarium of G. 

 honnemai with its long, narrow base, pointed for articulation, and 



