86 STRONG. 



barbules are absent as are also the barbicels. There is a very 

 heavy dark pigmentation of the dorsal halves of the barbules 

 and they are arranged so that they present their flattened surfaces 

 upward in a dorsal view. The dorsal margins rest upon the 

 median regions of the next more distal barbules and the ventral 

 margins are thereby hidden from view. The arrangement sug- 

 gests, somewhat, two rows of shingles overlapping each other on 

 either side of a median axis. 



On sectioning these barbules, I find an outer transparent layer 

 less than one micron in thickness, which encloses cell cavities 

 more or less closely packed with spherical granules of melanin 

 pigment less than one micron in diameter. The greater part of 

 this pigment is in the dorsal half of the barbule, the part visible 

 in a dorsal view of the feather. 



For several months I held strongly to the view that the thin trans- 

 parent layer produced the well-known interference colors of thin 

 plates, as was supposed by Altum ('54, '54a) and Brucke ('61). 

 The exceedingly uniform size and unusual shape of these pigment 

 granules seemed too significant, however, to warrant an unqualified 

 acceptance of the thin-plate hypothesis. Usually, melanin gran- 

 ules become more or less generally fused together into irregular 

 masses or are imbedded in the horn substance of the feather. In 

 the barbules which give metallic colors, however, the spherical 

 granules retain their individual form. Cross sections frequently 

 show them scattered about outside the barbule section and they 

 seem to have been simply packed in the cell cavities of the bar- 

 bules with little or no cementing substance to hold them together. 



In the feather germ, I find typical pigment cells or chroma- 

 tophores which produce typical rod-shaped granules. But after 

 these granules have passed into the cells composing the funda- 

 ments of the barbules which are to have metallic colors, they are 

 transformed into spherical granules. 



By a fortunate manipulation of apparatus, recently, I was able 

 to view barbules by strong reflected sunlight while using a Leitz 

 No. 7 objective. A beautiful pattern of gleaming spheres was 

 presented to my eye. Each pigment granule appeared to be 

 diffracting light and all the colors of the spectrum were visible 

 in the field. On the broad surface of the barbule, where the 



