ON THE PTERYLOSIS OF MESITES. 333 



anterior of the ventral powder-down patches to be presently described). 

 It recommences, however, a little lower down as a very narrow tract, 

 composed at first of only single feathers, but subsequently becoming 

 stronger and broader (though even here only two feathers broad), in which 

 condition it runs on, as the main inferior tract, to terminate near the vent. 



Strange to say, what must be considered the equivalent of the outer 

 pectoral branch of ordinary birds is here quite free throughout from the 

 main stem, with which it is not even united anteriorly, where it is sepa- 

 rated by the already-mentioned powder-down patch, whilst posteriorly 

 it runs parallel to, but quite free from, the main stem. 



The powder-down patches of Mesites resemble those of the Ardeidse, 

 of Leptosoma, and Podargus in their compactness, as well as in the de- 

 finiteness of their areas, as opposed to the more scattered and diffused 

 forms they present in Rhinochetus, Eurypyga, Crypturus, and other birds. 

 But in their exact distribution they differ materially from any of these. 



As already described by Mr. E. Bartlett, there are five pairs* of 

 powder-patches in Mesites. Of these two pairs are dorsal, two ventral, 

 and one lateral in position. All have the form of well-defined more or 

 less oval areas, covered by a dense mat of closely aggregated long powder- 

 down plumes. 



The most anterior pair is placed close to the median line, the patches 

 being only slightly separated from each other, at the commencement 

 (apparently) of the interscapular region and inclosed between the two 

 dorsal tracts, a little before these pass into their weaker posterior fork. 



The second dorsal patch is situated on the rump, close to and just out- 

 side the terminal part of the dorsal tract, between that and the posterior 

 termination of the lumbar tracts of each side. 



The third patch is inferior in position, lying just at the commencement of P. Z. S. 1882 

 the pectoral region, between the two halves of the inferior tract internally P 1 ' 

 and the anterior extremity of their separated pectoral branch externally. 



The fourth patch is longer and narrower than its fellows, lying on the 

 ventral region just outside the middle part of the inferior tract. 



The fifth (lateral) patch lies more or less transversely, in the neigh- 

 bourhood of the axilla, between the posterior ends of the outer pectoral 

 and humeral tracts. 



No Ardeine bird has any such lateral pair of patches ; and only Can- 

 croma has the anterior dorsal pair. These moreover lie outside, and not 

 inside, the dorsal tracts in that bird. On the other hand, the concen- 



* In the Ardeidae the number varies from one pair (Balceniceps) to four pairs (Can- 

 croma). Three is the most ordinary number. The presence of a single caecum in 

 Balceniceps (as fortunately demonstrated by a preparation mounted in the Museum of 

 the Eoyal College of Surgeons), together with these powder-down patches, renders its 

 Ardeine nature nearly certain, as already suggested by Mr. A. D. Bartlett (P. Z. S. 

 1861, p. 131). 



