chap, iil] VERTEBRA AND RIBS : BRADYPODID^E. 121 



On this evidence several comments suggest themselves. First 

 it should be noted that the Bradypodidae strikingly exemplify the 

 principle which Darwin has expressed, that forms which have an 

 exceptional structure often shew an exceptional frequency of 

 Variation. Among Mammals the Sloths are peculiar in having a 

 number of cervicals other than 7, and from the tables given it will 

 be seen that both the range and the frequency of numerical 

 Variation is in them very great, not only as regards the cervicals, 

 but as regards the vertebra? generally. 



As concerning the correlation between Variation in the several 

 regions, Welcker points out that his results go to shew that there 

 is such a relation, and that when the sacrum is far back, the ribs 

 also begin further back, or at least are less developed on the 

 cervicals. As he puts it. with a long trunk there is a long neck. 

 This is a very remarkable conclusion, and it must be admitted 

 that it is, to some extent, borne out by the additional cases given 

 above. The connexion, however, is very irregular. For instance, 

 the Cambridge specimen of Brady pus, though the 29 th is the 1st 

 sacral, has had cervical ribs of good size on the 9th vertebra, and 

 even has a small one on the 8th. But taking the whole list 

 together, Welcker's generalization agrees with the great majority 

 of cases. Expressed in the terms defined above, we may therefore 

 say that backward Homoeosis of the lumbar segments is generally, 

 though not quite always, correlated with backward Homoeosis of 

 the cervicals, and vice versa. 



It will be seen further that this Variation concerns every region 

 of the spine, and that even in the total number of prae-sacral 

 vertebrae there is a wide range of variation, viz. from 27 to 29 in 

 Bradypus (52 specimens) and from 30 to 34 in Cholozpus (20 

 specimens). Perhaps no domestic mammal shews a frequency 

 of variation in the fundamental number of segments com- 

 parable with this. In this connexion it may be observed that the 

 absolute number of dorso-lumbars in Clwl&pus (25 — 27) is ex- 

 ceptionally large amongst mammals ; but this is not the case in 

 Bradypus. 



If the case of Bradypus stood alone, some would of course 

 recognize the occurrence of cervical ribs on the 9th and 8th ver- 

 tebrae as an example of atavism, or return to the normal mam- 

 malian form with 7 cervicals. The occurrence of normal ribs on 

 the 7th in Gholoepus and the occasional presence of cervical ribs on 

 the 6th vertebra in this form, even reaching nearly to the sternum 

 as in Welcker's Leipzig case, obviate the discussion of this hypo- 

 thesis. 



We have, then, in the Bradypodidae an example of mammals in 

 which the vertebrae undergo great Variation as regards both their 

 total number and their regional distribution. As the tables shew 

 this is no trifling thing, concerning merely the number of the 

 caudal vertebrae, the detachment of epiphyses which may then be 



