652 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
and without perceiving both we get at nothing that is of much real 
use to us. 
A good history of Roman architecture has yet, however, to be 
written; and there are some difficulties in the way of writing it which 
are surprising if we consider only how many remains still exist and 
how much we know about Roman history and civilization in other 
directions. The difficulties are the paucity and architectural un- 
importance of existing remains that can even probably be assigned to 
periods before the establishment of the Empire, the frequent difficulty 
in dating with certainty the existing remains in Rome of later 
structures, and the entire absence, in most cases, of definite clues to 
the dates of provincial buildings. The first two difficulties are largely 
due to the building activity of the first two centuries of our era, 
which led, with considerable aid from fires, to the repeated partial or 
complete reconstruction of almost all important buildings in Rome. 
The uncertainty of the chronology of Roman architecture was 
somewhat violently illustrated only a few years ago. In 1894 M. 
Chedanne, a young Frenchman, seizing a favorable opportunity, was 
able to show that the rotunda of the Pantheon is not the building 
erected by Agrippa in B. C. 27, as had generally been supposed, but 
almost certainly a structure of the time of the Emperor Hadrian; 
quite certainly not earlier, for all the bricks which he extracted at 
random from various parts bore stamps known to be of that reign.* 
This discovery removed the principal witness to the extraordinary 
fact, hitherto always assumed, and often asserted in so many words, 
that the Imperial Roman system of brick and concrete construction 
sprang suddenly into existence fully developed in the reign of Augus- 
tus, and endured, virtually unchanged, for nearly three and a half 
centuries. M. Choisy himself, in a short chapter? labeled—perhaps 
in irony—* historical,’ spent much language in enlarging on the 
unprecedented and surprising nature of the thing; but he insisted 
nevertheless on the sudden rise of the system, its almost as sudden 
abandonment, and the absence of growth or development between. 
M. Choisy, like others, evidently compared the Pantheon with the 
writings of Vitruvius, who must have published them at just about 
the time it was being built, yet was obviously entirely ignorant of 
any methods of building nearly so advanced; and instead of finding 
what now appears to be the obvious solution of the puzzle—namely 
that the date assigned to the Pantheon was a mistake—M. Choisy 
accepted the date as others had done and arrived at the only conclu- 
sion then possible—namely that the incredible had really happened, 
that the system on which the Pantheon is built had just sprung sud- 
denly into existence, and that Vitruvius was a poor old fellow far 
aR, I. B. A. Journal, 1895, 176-177. ‘VT’art batir, Part 3, chap. 1. 
