180 NATURAL HISTORY OF BIRDS. 



"good and sufficient to unite the Steganopodal group together, and at the same time 

 mark it off from all other groups of birds." 



Another thing is, that Mivart has shown that the four supergenera, included in 

 brackets above, are more intimately related inter se than to the two other ones. 

 These two, on the other hand, chiefly agree to differ from the former four in negative 

 points, and hence their exclusion from these does not indicate any particular mutual 

 intimacy. On the contrary, the tropic-birds and the frigate-birds are as different 

 between themselves as each of them is from the rest. We therefore propose to dis- 

 member the order in three superfamilies, Pelecanoideae, Fregatoidea3, and Phaethon- 

 toideae, an arrangement which is essentially the same as that proposed by Professor 

 Brandt forty years ago. 



This arrangement needs a short explanation. There will be found, later on, a few 

 more details concerning the peculiar arrangement of the neck vertebrae of the first- 

 mentioned superfamily. In the last two the neck is normal, and, consequently, they 

 have not developed as off-shots from the stem of the Pelecanoideae. The extreme 

 specialization of Fregata in regard to the thigh muscles, A+, can, therefore, not be 

 derived from the Pelecanoideae, notwithstanding that the myological formula of the 

 latter, AX+, otherwise would allow of such an interpretation. On the other hand, it 

 is even more plain that the myological formula, AXY -r, of the tropic-bird cannot 

 directly or indirectly be derived from the pelicans or the frigate bird, nor, indeed, 

 the latter two from the former. We are, consequently, compelled to assume a 

 common ancestor with normal arrangement of the cervical vertebrae and a myological 

 formula consisting of, at least, AXY+. 



We have occasionally had opportunity to hear people ridicule the stress laid upon 

 the presence or absence of such a trifling thing as a small muscle of the leg seems to 

 be. In some instances the presence or absence, considered alone, throws no light upon 

 the manner in which two forms have developed, and in other cases it seems to the 

 superficial observer to have no systematic importance, for instance, when a species has 

 a certain muscle which is wanting in a closely allied form of the same genus ; but 

 even then it is of considerable interest, since it shows that the latter has developed 

 out of the former, and not vice versa. The above example, however, derived from 

 the present order, should convince even the most superficial observer that there are 

 cases in which these tiny muscular slips play a most important role. 



We have discussed the distinctness of the three groups here proposed only on the 

 basis off a few characters, since want of space prevents us from going further into 

 details ; but in order to show that the differences are rather deep-rooted, it will be suf- 

 ficient to remind one of the fact that they are apparently not due to direct teleological 

 causes. In all three groups there are excellent flyers, with long wings ; but one of 

 them also comprises rather short-winged divers. Similarity in habits and manner 

 of life may account for the external and supei'ficial resemblance between a gannet and 

 a tropic-bird, but we know of no difference in their habits sufficient to explain the 

 anatomical diversities alluded to above. 



In addition to the characters common to all the members of the group, as given 

 at the beginning of this chapter, the double condition of the pectoral muscle is here 

 described in Professor Garrod's words : 



" The great pectoral muscle is composed of two independent layers, a superficial 

 large one, arising from the inferior border of the sternum, its carina, and from the 

 outer border of the furcula ; and a deep one from the upper two thirds of the deeper 



