Ill] OF FIELDFARES 265 



domed building, a peristyle/ as it were, covered 

 with tiles or net, is constructed, in which several 

 thousands of fieldfares and blackbirds may be en- 

 2 closed, though some people add to them other birds 

 as well, which, when fattened, fetch a good price, 

 such as ortolans^ and quails. Into this building 

 water should be brought by means of a pipe, and 

 it had better then flow slowly along narrow troughs 

 such as can easily be cleaned out (for if the water 

 spreads over a large area it is more easily fouled 

 and it is not so good to drink), and the over- 

 flow ' from these should be taken off by a pipe, lest 



of aviary from which birds are taken {sumuntur) to the market, 

 not that of Lucullus, in which they are eaten {sumuntur — 

 consuviuntur). 



^ Ut peristylum. Keil thinks the text sound here, Schneider 

 proposes aut^ and remarks that a testudo^ the essential part of 

 which was its dome, would need neither tiles nor net. 



^ Miliariae. Cf. Varro, L. L., v, 11 : Ficedulae et miliariae 

 a cibo quod alterae fico alterae milio fiant pingues. The Greek 

 word KtyxpoQ means "millet," and Ktyxplg seems to be the 

 "ortolan." 



' Caduca. Keil quotes here Fronto (De Aquis, ii, 94), to 

 show that aqua caduca was a technical phrase to indicate the 

 overflow from a tank or the droppings from a pipe, and goes 

 on to remark : Ex quihus apud Varronem dilucidafit verhorum 

 coniunctlOy in qua interpretes haeserunt. But I cannot see how 

 this gets over the difficulty perceived by Ursinus and Schneider. 



The water goes in by a pipe, is distributed along several 

 small runnels or troughs (canales) which converge at the 

 further end of the aviary to another pipe by which it flows out 

 (exit) — so that there is a continuous flow of fresh and clean 

 water. This seems to be the obvious meaning. But then 

 what is to be done with the words quae abundat? How 



