Summary.Thus, from a study of the cranial osteology, the pterylosis, the geographical distribution, the plumage pattern, the bower‐building habit, and the eggs, I am unable to find a single feature which warrants the inclusion of the Ptilonorhynchidse with the Paradiseidae. The structure and the colour‐pattern of Loria ally it unmistakably with the latter group.I have not been able to investigate the structure of Prionodura, Sericulus, Cnemophilus, æluroedus, or Loboparadisea. The first two are quite clearly Bower‐Birds, since they build bowers, and have the typical coloration in both sexes. Cnemophilus has, likewise, the characteristic yellow and black male and the brown and grey female of some of the others, and has no superficial characters that I can discover to suggest any relationship with the Paradiseidae. ælurcedus, with its green and mottled markings, and very stout bill, has little in the way of superficial characters to ally it with either, and I do not know on what grounds it was first placed among the Ptilonorhynchidae, but until its relationships can be established by dissection, we must presumably keep it there. Loboparadisea is another rather problematical case; as already indicated, it is not typical of either group, but in view of the greater uniformity existing among the Ptilonorhynchidae as compared with the Paradiseidae, I regard it as an aberrant member of the latter family.With regard to relationships with other groups than the Birds of Paradise, Gadow (1888) has suggested affinity with the Wattled Crow (Calleeas) of New Zealand. But this form does not show a single common feature in the skull, which possesses very individual structure both of the palate and the lachrymal. Likewise (contrary to the statement of Nitszch) the pterykrais is dissimilar, and there is no apterion in the dorsal tract. The New Zealand Thrush, Turnagra, is another form whose relationship to the Bower‐Birds has been suggested. But from his investigation of the anatomy Gadow (1888) was quite unable to confirm this.As far as I can see, the Ptilonorhynchidae constitute a singularly complete and isolated family of the acromyodian passerine birds and show no special relationship to any other, being sharply marked off by the structure of the skull, the colour‐pattern, and the bower‐building habit; from the evidence I have endeavoured to bring forward I exclude Loria and Loboparadisea, and re‐consiitute the Ptilonorynchidss, as possessing family rank, and consisting of the following genera:–ælurædus. Ptilonorhynchus.Amblyornis. Scenopætes.CMamydera. Sericvlus.Gnemophilus. Xanthomelus.Prionodura.