l8o H. H. NEWMAN. 



I interpret the figures for both double bands and double scutes 

 as I formerly interpreted similar facts brought out by a study of 

 supernumerary scutes in the carapace of tortoises. These super- 

 numerary scutes were viewed as atavistic variations or vestiges 

 of a condition largely outgrown by the species. These anomalies 

 occur most frequently at the posterior end of the carapace and 

 progressively less frequently as one goes toward the anterior end, 

 except that there is a slight increase just at the anterior end. 

 The hypothesis was ventured that the more frequent the regional 

 occurrence of an archaic character the more recent has been the 

 racial suppression of these characters in that region. There 

 has evidently been an antero-posterior orthogenetic loss of scutes 

 in the tortoise carapace beginning at or near the anterior end 

 and ending with the posterior end. 



Now in the armadillo I believe that the more generalized and 

 less regular parts of the carapace, the scapular and pelvic shield, 

 represent phylogenetically an older condition than does the 

 banded region. In support of this contention I cite the fact that 

 the extinct giant armadillo, Glyptodon, had a solid non-banded 

 carapace. The first banding began in one or two bands at about 

 the middle of the carapace, where bands 5 and 6 are, and pro- 

 ceeded posteriorly and anteriorly. Progress anteriorly in the 

 direction of more bands is probably occurring now, as may be 

 judged by two facts, first that it is not uncommon to find a part 

 of the last scute row of the scapular shield separated as a band, 

 and second that there are not infrequently local fusions between 

 parts of the first band and the scapular shield. On this hypothesis 

 band I is the newest band phylogenetically and shows more re- 

 semblance to the scapular shield than do other bands, both in 

 band doubling and in scute doubling. Band 2 is much less 

 affected with these anomalies than is band I, but more so than 

 any other band. Moreover, bands 5 and 6, which are phylo- 

 genetically the oldest bands, show the fewest recurrences of 

 irregular conditions. Scute doublings are viewed as incipient 

 or vestigial band doublings and we find many more of these 

 vestigial anomalies in the banded region than fully developed 

 ones. 



In the tortoises the orthogenetic progress in carapace simpli- 



