Avirvltine in the Dcaj^ nj Ancient Rome. 327 
on a good and sufficient water supply, and prefers that it 
should be a slow stream running continually through shallow 
and narrow troughs, so that these may be Jrequently cleaned; 
the overflow should be carried away by a pipe to prevent 
the formation of mud. Entrance was to be elTected by means 
oi a '■ coclia " or rotating cage, similar, I take i^, to the exit 
arrangements at certain parts of our London Zoo. The win- 
dows were to be few, lest the hapless victims, seeing Ijirds 
and trees outside should pine, but sufficient light allowed to 
enable the birds to find food and water, and to locate their 
perches. The doors and windows were to be surrounded witV 
smooth plaster to exclude mice and other undesirable aliens. 
Varro evidently appreciated the fact that softliills cannot live 
on grain, for he recommenderl, as food, cakes tnade of figs and 
barley, well mixed together and rolled. Besides perches and 
ladder -like erections for tl:e birds to rest on, he sugge-;ted the 
walls should be furnished with shelves and recesses. The 
object of these was (1) for the birds to hide in and (2) that 
aviarium (=aviary attendant) might deposit the Iiodi.'s of 
thn liirdi that died a natural death. It is thus comi'orting to 
know that the unexpected death of one's best birds is not 
an altojjether modern exasperation. The birds remained in the 
main aviary until a public feast or other excuse for eating 
arose, when they were transferred to the Seclusorium, which 
was rather lighter and the aviarium was admonished to kill 
those appointed to die" in secret and away from the others, 
lest these, seeing it, should mope, and die at a time incon- 
venienl to the seller." This last sentence rather suggests 
that at times the feast had either to 1>e hurried on or in- 
deflnitely postponed according as whether the Ijirds weiv ill or 
well, although \'arro does not specifically say so. 
Varro, in writing to Axius makes a few naive remarks 
on the habits of Fieldfares, and infoi-ms him that they do not 
nest on the ground like other immigrant birds such as th(! 
Stork, nor under the roof like Swallows. He i^laces the 
period of immigration at the autumnal eijuinox and the return 
of the 1)irds at or about the spring equinox. fie does not 
venture or. their ultimate destination but remarks on the 
migration of the common Quail and Turtle-IJove. Finally 
Varro deals with th(>. question of C s. d. The balance-sheet 
