Xll 
the oil-gland. Anteriorly the two divisions of the spinal tract are connected by scattered 
feathering with the humeral tract. The ventral tract is double from close to its point of origin 
at the mandibular symphysis ; the two tracts are at first very narrow, and separated from each 
other and from the sides of the head by spaces without any feathering : on the pectoral region 
each ventral tract is of uniform width, four or five feathers wide ; below the axilla each tract gives 
off a closely feathered lateral branch to the hypopteron. The pectoral tract of each side dwindles 
to a single row of feathers before ending just in front of the cloaca. 
“ The pterylosis of Merops does not differ in any important respects from that of Coracias , 
which it appears most closely to resemble ; it differs from Momotus and Todus in having a double 
spinal tract, and from G-albula in having no inner branch to the pectoral tract running along 
the clavicles. 
“OSTEOLOGY. 
“ The characters of the skull of Merops have been dwelt upon by Prof. Huxley, in his well- 
known paper on the Classification of Birds. The skull is desmognatkous, and the family is placed 
in the group Coccygomorpk.se, in close proximity to the Alcedinidse, Bucerotidge, Hpupidse, 
Momotidse, and Coraciidse. In the same paper the author refers to the shape of the sternum in 
the Coccygomorpkse, which is stated to present usually two notches on each side, and to be devoid 
of a bifurcated manubrial process, the one exception to the last statement being Merops. I find, 
however, that Trogon ( T. reiriliardti ) has a distinctly bifurcate manubrial process, though 
not perhaps so well marked as in Merops. The characters of the skull of the former genus, 
which has been shown by Mr. Porbes to be sckizognathous and not desmognatkous, would appear 
to necessitate its removal altogether from Huxley’s group Coccvgomorphse, though it undoubtedly 
agrees in its visceral and muscular anatomy, as well as in external characters, with the other 
families of this order, and presents affinities to Merops, as well as to the Celeomorphse and 
Coracomorphae, by the possession of a forked manubrium sterni. 
“ The sternum of Merops is, as stated by Prof. Huxley, provided with two lateral notches, of 
which the outer one on each side is the deepest ; in Coracias, Trogon, Tleroglossm, Bucco, Bacelo , 
and Yunx the sternum is closely similar; in Momotus the notches are converted by continuous 
growth into foramina ; in TJgyupa and Buceros there is only a single pair of notches. 
“ According to Prof. Huxley the clavicles of the Coccygomorphse are without any process 
developed backwards from the summit of their e symphysis.’ This statement is true, in so far 
that there is no junction between such a process and the sternum as exists in many Passeres, not 
to mention other orders of birds ; but the process it, self exists, in a rudimentary form, in many of 
the Coceygomorphse. In Coracias and Trogon there is such a process developed just at the 
junction of the two clavicles, and there is the faintest trace of a similar structure in Merops. On 
the other hand, Upupa, Yunx, Bucco , Momotus, &c. have no median process of the furcula. 
