1777 Neo had several distinct singing periods and each had a different basic theme; the tendency being, in any given period, to lay emphasis on one particular theme. Again it was observed, as I listened from about 10 feet away, that full song is ushered in by a number of soft introductory notes which can not be heard more than a few feet away and that there are, introduced into the song, similar soft "grace" notes not heard say 20 feet away or so. All this checks with observations on Brownie. Rhody, at noon, came out of the brush on the west lot on call but would not come over the fence and follow for his mouse, preferring to warm his back in a sunny place. (Why did he come to my call?). The Anna Hummer astonishes me by going into a blue-gorget phase for 10 minutes! 1:50 P.M. The humming-bird patient was taking a bath in his cage at 1:15: a long affair, as he was trying to wet himself all over-apparently, while sitting on a perch a little too high above his bath dish. I helped him by spraying a mist on his back with an atomizer. He co- operated fully and became "black". I glanced at him occasionally as he dried himself and was amazed to catch a glint of clear ceru- lean blue from his gorget: a color which does not appear in this bird normally. In certain lights and at certain angles the rosy crimson of gorget and crown often appears to be overlaid by a sort of halo or fluorescence like the bloom on a mineral lubricating oil. This bloom seems almost to be detached from the feather itself and to be in space above it. (Not that I think it is, but that is the nearest I can come to describing the effect given). I have no color charts to which I can refer, so the following colors are as I see them. Normally the colors appearing in the gorget, crown and small patches to the rear of the eye are, of course variable, depending upon angle at which the light strikes, is reflected, viewed, etc. I see crimson, rosy crimson, scarlet, magenta, gold, bronze, green running from emerald through through various mixtures with yellow to brilliant clear yellow; but never any blue other than the bloom referred to. Now, as I watched, this "heavenly" blue occupied more and more of the area of gorget and head and was streaked with magenta or rosy crimson. It was overlaid by no other color. Some areas were entirely of this blue: a body color--not a bloom--darker than the bloom, and still metallic. I was so amazed at this totally unex- pected phenomenon (apparition) that I called my Filipino boy, Julio, (who incidentally waits on this bird "by inches") and asked him what he saw. He said he saw blue and the regular colors and that the colors "go with each other". (He has some eye for color combinations. Once when I had planted a certain fuchsia by the front steps in order to get the effect of its bloom against the sandstone, without my tell- ing him what I was doing he volunteered the opinion that the flower "goes with the house"). As the bird became drier the blue was gradually replaced by the regular colors of the dry bird. In about 15 minutes no trace of the blue was left--I had put him in the direct sun close to the warm wall of the cloister. Without attempting a complete analysis of the mechanism of this phenomenon--a too ambitious attempt to be made at the moment of its discovery, if indeed a complete explanation be possible, it is clear that the color change was due to the feathers' being wet. The metallic colors of gorget and crown are structural colors overlying whatever colors, due to pigments, may be the body colors of the feathers themselves. These metallic colors are primarily due to interference effects caused by surface striation, scales, plates; i.e. minute physical