1511 back with it to the cage, followed by him until he caught sight of the group sitting there, when he retreated. As he did not reap- pear, I went back to look for him and found him, as expected, back in the shop-yard, patiently waiting by the mouse cage, look- ing up at it. He came out obediently when called and I returned by another route, followed by him. By this route he would not have to ap- proach the group from the front, but come up from the rear. This he did, but was careful to make a detour around them through the bushes, taking the mouse with considerable reluctance from about 15 feet in front of them and retreating quickly. Rather unexpectedly, in about a half hour, he followed Julio to the yard for a second mouse, which he also ate. Still later he returned to his sunning place by the loquat, but being careful to keep as much out of sight as possible while going there. August 30th. Rhody's moults. About 9:30 A.M. Rhody approached the cage and I led him to the yard for his first mouse of the day. One of his tail feathers was dragging with tip on the ground, but fell out as I watched. This feather, it happens, is the left-side mate of the first tail feather shed this year, found by me May 12th. They are mir- or images of each other. Thus it would appear that the moul t of Rhody's rectrices has already covered a period of 3½ months, and since it required about two months for him to replace completely the seven tail feathers lost by him through accident on October 26th. 1936, it would seem that at least 7½ months will be requir- ed for his tail moul to be completed--perhaps longer, as I am not sure that this rectrices have all been shed as yet. There are still several that have not yet reached their full growth. "Hunger Marks" These two feathers were carefully examined for "hunger marks" without finding any. They were also examined for "water marks" (assuming that I understood Dr. Grinnell's description of these verbally on the 23rd inst. and can identify them). I counted 60 on each, plus or minus 2 or 3, allowing a mar- gin for possible error. (Note how this number fits in with the number of days taken in completely replacing the lost feathers of October 1936). What I took to be "water marks" were transverse bands across the vanes and therhachis and the calamus. On the vanes they appear as rose-purple bands, when the feather is held at a certain angle, in sharp contrast to the blue-green background; this being an optical effect probably due to the fact that the bands occupy slight depressions across the vanes and the light from them (that enters the eye) has been reflected at an angle differing from that at which the light from the rest of the feather (that enters the eye) has been reflected. It appears to be a diffraction and inter- ference effect. On the rhachis the marks appear as shallow grooves; on the calamus as whitish bands. The figure 60 represents the sum of the bands on the vane plus those on the calamus beyond where the vane ends. These marks are not what I have called "hunger marks" after Pycraft (Encyc. Brit., llth. Ed., Art. FEATHER). Pycraft says: "The growth of the feathers is, however, certain- ly affected by the general health of the bird, mal-nutrition c causing the appearance of peculiar transverse V-shaped grooves, at more or less regular intervals, along the whole length of the feath- er. These are known as "hunger-marks", a name given by falconers, to whom this defect was well known". They are not at all like the "water marks" in appearance nor do they appear in the old, worn