128 REPORT—1852, 
This kind of chair and sleeper is applicable to iines already laid, and to any pat- 
tern of rail, and when it is necessary to relay the line, or renew the sleepers. 
In concluding his paper, Mr. Godwin suggested the possibility of constructing a 
railroad in such amanner that the engine and carriages should never rest on a joint, 
but roll over a continuous and unbroken line of rails. He thought that this might 
be effected by rolling the rail in two halves, and riveting them together in such a way 
that the engine would never bear wholly on a joint ; and by adopting the wedge-fast- 
ening alluded to. above, the strain on the rivets would be relieved and not likely to 
shake loose ; indeed, the section of the rail might be so contrived that the action of 
the wedge in keeping the rail tight in the chair may also tend to force the half rails 
together, and probably render rivets altogether unnecessary; of course it would be 
necessary to provide for expansion and contraction, which could easily be effected. 

On a Dynamometric Machine for Measuring the Strength of Textile Fabrics 
and other Substances. By M. PERREAUX. 
On Telegraphic Communication between Great Britain and Ireland, by the 
Mull of Cantyre. By W. J. Macquorn Rankine, C.L., and Joun 
Tuomson, CLE. 
The authors recommend the construction of a line of electric telegraph between 
Great Britain and Ireland, crossing the North Channel from the Mull of Cantyre to 
Tor Point, chiefly on the following grounds :— 
1. It involves the construction of a much less length of submarine telegraph than 
any other line, the distance across the channel in the line proposed being only thirteen 
miles, while that from Portpatrick to Donaghadee is twenty-two miles; from five to 
five and a half additional miles of submarine telegraph would be required to cross 
small arms of the sea; but this would be in small detached portions, easily laid and 
repaired, and would make the total length of submarine telegraph only about eigh- 
teen miles. 
2. It is the most secure of all sites for an electric telegraph between Britain and 
Ireland, for no vessel ever casts anchor in the proposed line. 
3. Besides these national advantages, it has the local advantage of connecting the 
North-East of Ireland directly with the ports on the Clyde. 
As it may be considered necessary for local purposes that the electric telegraph 
should be carried as far north as Larne in Ireland, and as far westward as Dumbarton 
and Greenock on either bank of the Clyde, Larne may be looked upon as the Irish 
terminus of this scheme, and either Dumbarton or Greenock as the Scottish terminus. 
The line from Dumbarton would require 106 miles of land telegraph, the alternative 
line from Greenock 93 miles, to complete the communication. 
The authors consider the security of this line of telegraph to be an advantage, in a 
national point of view, sufficiently great to warrant its execution, even were the lines 
by Portpatrick and Holyhead in full operation. 
Remarks on the Mechanical Process for Cooling Air in Tropical Climates 
proposed by Prof. C. Piazzi Smyth. By W. J.M. Rankine, C.E., F.RSE. 
The most improved form of the apparatus proposed by Prof. Smyth consists,—1. of 
a compressing pump, by which the air is to be forced into, 2. a refrigerator, consist- 
ing of a long tube, or a series of tubes, exposed to a stream of water, in which the 
air will be deprived of the heat generated by the compression, and from which it will 
escape into, 3. an expansion cylinder, in which the air will at once become cooled 
by expansion to an extent nearly, but not quite, equal to that of the original heating 
by compression, and will propel a piston, to assist in working the compressing pump. 
The air will be delivered from this expansion cylinder into the building to be venti- 
lated. The principal resistance to be overcome in this improved machine will be the 
friction. The author gives formule and rules for calculating the dimensions of the 
parts of this machine, and the power required to work it, supposing the friction to be 

