TfiE BRITANNIA AND CONWAY TUBULAR BRIDGES. 419 
facts, and having always had an antipathy to all forms of 
trussed beams, Mr. Fairbairn appears to have believed from the 
first that a simple tubular girder, without auxiliary chains, but 
of the form first suggested by Mr. Stephenson, would best 
meet all the requirements of the case. In this opinion neither 
Mr. Stephenson nor Mr. Hodgkinson concurred till a much 
later period ; the former was determined “ to have two strings 
to his bow ” (both chain and girder), and the latter was scep- 
tical as to the resistance of the tubes to compression. It is a 
curious fact in proof of this, that the Conway Towers, com- 
menced as late as May, 1846, were built with the necessary 
provisions for the reception of the chains as originally de- 
signed. 
II. The experimented Inquiry . — The general principles of 
girder bridges were already ascertained ; the form of greatest 
strength in cast-iron had been determined by Mr. Hodgkinson ; 
but of wrought non much less was known. The strength and 
proper distribution of this material was, therefore, the subject 
which Mr. Fairbairn set himself to investigate in a series of 
experiments. He at once proceeded (July 1845), abandoning 
all reference to suspension-chains, to test the actual resisting 
powers and mode of fracture of various kinds of tubes. Cylin- 
drical tubes, as the simplest, were first tested ; then elliptical 
tubes, specially recommended by Mr. Stephenson ; and lastly, 
rectangular tubes. The rectangular tubes were found to be 
decidedly the strongest, next to these the elliptical, and weakest 
the cylindrical. Thus the first of the distinctive principles of 
the bridges — the rectangular form, was at once deduced from the 
experiments ; and the other was soon to follow. Almost all the 
tubes had given way by the crumphng or buckling of the 
upper part, as shown in PI. XXIII., fig. 3. The resistance of thin 
plates to compression was evidently very inferior to their resist- 
ance to tension. It was necessary to strengthen the top, and 
this Mr. Fairbairn effected by means of a cellidar fin. A tube 
was constructed of an elhpitical form and with a triangular cell 
at the top (PL XXIII., fig. 1), and the result was so satisfactory 
that a rectangular tube with two cylindrical corrugations or 
cells was at once planned (fig-. 2). This tube (which obtained 
the name of the spectacle tube) was experimented upon Octo- 
ber 14, 1845, and first combined the two novel features of the 
bridges. From this period every difficulty in the design of the 
bridges disappeared. 
In March, 1846, the first definite calculations of the areas of 
metal in the Conway tubes were made, and a model was con- 
structed of l-6th the real dimensions, which was submitted to 
experiment with a gradual increase of area in the bottom until 
the resistances to tension and compression were precisely 
