154 SECTIONAL ADDRESSES. 
There were no large clearing or settling tanks within the city, only 
comparatively small reservoirs (castella) from which distribution was made 
by lead: pipes ; and this is the case with the modern aqueducts also, so 
abundant is the supply. 
From the study of the aqueducts we should pass naturally to that of the 
buildings which they supplied. The great imperial thermae, such as those 
of Caracalla, were naturally among the most important of these ; and in 
this case, as in others, the water supply of the city as a whole was increased 
in order to have enough water for the special aqueduct, a branch of the 
Marcia (to which he added a spring called the Fons Antoninianus, after 
his own name) by which they were supplied. 
But this might lead us on into a general survey of Roman architecture, 
which would be far too vast a subject. It may perhaps be opportune to 
ask, in conclusion, who were the practical men who carried out these great 
works ? 
The question who were the persons responsible for the development 
of Roman engineering has been asked and answered by Rivoira in his 
Architettura Romana, soon to appear in an English translation, the proofs 
of which I have been allowed to see. He points out that the architects 
of the great Roman state buildings of the Imperial period were forbidden 
to inscribe their names upon them, but rightly maintains that they must 
not be assumed to have been Greeks, in that prejudice and passion for 
things Greek against which he raises a very necessary protest. This pre- 
judice leads those who are swayed by it to ‘look at the architect solely 
in his character of artist and exponent of aesthetics, forgetting the technical 
and engineering sides of his activity.’ Rivoira emphasises the fact, as 
does Montauzan,°° that we learn from Vitruvius what class of men 
these state architects were—military engineers, who were at the same 
time civil architects. They were, it would seem, all Roman citizens, and 
for the most part Italians, not Greeks, as their names imply ; and it is 
to them that we owe the planning of such a garrison town as Aosta, and 
the development of Roman vaulted buildings, the ancestors of the great 
vaulted architecture of the Byzantine, Lombard, and Romanesque periods. 
The Roman engineers had at their disposal comparatively simple and 
primitive instruments—the best work on the subject I know is M. Germain 
de Montauzan’s Essai, for it is written by a practical man. 
The groma, their chief land-surveying instrument, has recently been 
reconstructed from fragments found at Pompeii by Sig. Matteo della 
Corte,*! and there is a reproduction of it made by Col.-Sir H. G. Lyons 
in the Science Museum at South Kensington. It can only be employed for 
measuring horizontal angles; for measuring vertical angles they had to 
proceed by slow degrees, sighting through a dioptra, which measured 
angles of both kinds, being combined with a water-level, or using a chorobate, 
a somewhat cumbrous form of water-level, much like a dumpy level of the 
present day, and preferred by Vitruvius for levelling aqueducts, probably 
because it involved no sighting ; for we must remember that they had not 
the assistance of the knowledge of the optical properties of glass. 
Their systems of calculation, too, which involved the use of the abacus, 
must have been complicated, slow, and inconvenient. 
80 Hssai sur la science et Vart de V Ingénieur (Paris, 1909), 114 sqq- 
3! Monumenti dei Lincet xxviii. (1922), 5 sqq. 
