258 Mr. R. Swinhoe on Formosan Ornithology. 



found several rows of perforations made by this bird. The birds 

 were flying in and out of them in great numbers, so we stopped 

 to examine them. Most of the holes were out of arm's reach ; 

 and as the bank was very steep, and composed of loose mud, we 

 had great difficulty in establishing a footing. We managed, 

 however, after much trouble, to insert our arms into several of 

 them. The holes were in all stages of progress, some only just 

 begun, others scarcely a foot deep ; in some the eggs were hard- 

 set, in others quite freshly laid. The holes ran into the bank 

 with only a slight inclination from the horizontal. In all in- 

 stances they turned a little to the right, extending in depth to 

 about 2 feet, — their diameter being from 2 to 3 inches, which is 

 enlarged to a cavity about 6 or 8 inches broad at the bottom. 

 In its cup-shaped base was placed the nest, composed of light 

 straw and dried grasses and lined with feathers. One nest, 

 however, had no feathers ; but as it had no eggs, I concluded it 

 was unfinished. The eggs in every case were only three in 

 number, of a pinkish white, without spot or stain. On our dis- 

 turbing the birds they rushed in consternation from their nesting- 

 site, and after flying about low in the air at some distance in 

 great agitation, they would meet together for some seconds as 

 if in consultation. They would then again hurry off in different 

 directions, and again meet. Finding we were in no hurry to 

 leave their ground, they began to scatter and soar away to a con- 

 siderable height. As soon, however, as we withdrew for a space, 

 they returned, many diving at once into their burrows, others rush- 

 ing backwards and forwards close past the holes, as if bewildered 

 and afraid to enter. They were so numerous, and looked so 

 small in the bright quivering light of a hot Foi'mosan day, that 

 they seemed to me at times more like Dragon-flies than birds. 



cJ, shot Oct. 10, 1861. Heart ^ in. by ^q. Liver, right lobe 

 ■^ in., left ^. CEsophagus ^ in. wide ; proventriculus -f^ in. 

 by ^. Gizzard rounded, with a small lower protrusion, flattened, 

 with strong tendons, h in. long by -^q, and ^ deep ; epithelium 

 thin, with broad longitudinal furrows. Testes transparent and 

 oval, Y^Q in. long. Caeca, 1 inch from anus, ^ in. long, and thin. 

 Intestine 5 ^q in. long, varying in thickness from ^ to -j-^ . 



I find, on closely inspecting the Sand-Martins procured near 



