Pliny opens the third letter of his collection, to Caninius Rufus, with a series of questions about the latter’s luxurious villa in Comum – I mark the clauses I am most interested in (1.3.1):(1)
Quid agit Comum, tuae meaeque deliciae? quid suburbanum amoenissimum, quid illa porticus uerna semper, quid platanon opacissimus, quid euripus uiridis et gemmeus, quid subiectus et seruiens lacus, quid illa mollis et tamen solida gestatio, quid balineum illud quod plurimus sol implet et circumit, quid triclinia illa popularia illa paucorum, quid cubicula diurna nocturna?
In The Letters of Pliny: a Social and Economic Commentary (Oxford, 1966, ad loc.), A. N. Sherwin-White give three parallels to Epistle 2.17, in which Pliny describes his own Laurentine villa: there are elegantly varied references to rooms catching the sun from different directions at different times of day in § 8, 13, and 23. As his “Cf.” implies, Caninius’ villa must be designed to do the same.
It seems to me that one sentence of Pliny’s description of his other villa at Tifernum (Ep. 5.6), is even more pertinent, at least grammatically and textually (5.6.31):
Hac [porticu] adeuntur diaetae duae, quarum in altera cubicula quattuor, altera tria ut circumit sol aut sole utuntur aut umbra.
The architecture is not quite parallel: here we have a suite of rooms designed so that some will be sunny, others shady, all day long. Nevertheless, in comparing the two passages, it seems obvious (at least to me) that 1.3.1 should be emended to read quid balineum illud quod plurimus sol implet ut circumit? I think we can conclude that Caninius’ bathhouse has windows facing southeast, south, and southwest, to catch the sun as it goes around. Whether the bathhouse itself was curved like Pliny’s cubiculum in hapsidi curvatum (2.17.8), or more angular, I cannot tell.
(1) Text and variants are quoted from Mynors’ Oxford Classical Text (1963). There are no variants pertinent to my argument. As for impertinent variants, I do wonder why editors print platanon in Roman letters, when one manuscript family (α) gives ΠΛΑΤΑΝΩΝ (with a grave accent over the omega). Surely scribes are more likely to have transliterated from Greek to Latin than the other way around.