ON THE HISTORY OF HYDRAULICS, &c. 277 



little further progress until the revival of letters. The Romans had water 

 mills in the time of Julius Caesar, which are described by Vitravius ; and 

 it appears that their aqueducts were well built, and their waterpipes well 

 arranged. Pipes of lead were, however, less frequent than at present, 

 from an apprehension of the poisonous quality of the metal, which was 

 not wholly without foundation.* Some say that the ancients had no 

 chimnies, but whatever may be the authorities, the opinion is extremely 

 improbable. 



It was in the middle ages that navigable canals began to be considerably 

 multiplied, first in China, and afterwards in other parts of the world. 

 The canal from the Trent to the Witham, which is the oldest in England, 

 is said to have been dug in 1134. The date of the earliest windmills has 

 been referred to the year 1299. The invention of gunpowder possesses 

 perhaps an equal claim with the art of printing, to the honour of being 

 considered as constituting the most marked feature that distinguishes the 

 character of ancient from that of modern times ; its introduction must 

 necessarily have tended to produce material alterations, and perhaps im- 

 provements, in the habits of nations and of individuals. It is said to have 

 been known long since to the Chinese, and our countryman, Roger Bacon, 

 was evidently acquainted with its properties ; but it was not actually em- 

 ployed in Europe or in its neighbourhood till about the year 1330 ; and the 

 earliest artillery appears to have been that which was used by the Moors, 

 at the siege of Algesiras, in 1334. King Edward had four pieces of cannon 

 at the memorable battle of Cressy, in 1346. 



About the year 1600, Galileo made the important discovery of the effects 

 of the weight and pressure of the atmosphere,t in the operation of suction, 

 and in various other phenomena. Before his time, it was generally sup- 

 posed that water was raised by a sucking pump, on account of the im- 

 possibility of the existence of a vacuum : if, however, a vacuum had been 

 impossible in nature, the water would have followed the piston to all 

 heights, however great, but Galileo found that the height of its ascent was 

 limited to about 34 feet, and concluded that the weight of a column of 

 this height was the measure of the magnitude of the atmospherical pres- 

 sure. His pupil Torricelli afterwards confirmed the explanation, by 

 showing that a column of mercury was only supported when its weight 

 was equal to that of a column of water standing on the same base ; hence 

 the vacuum obtained by means of mercury is often called the Torricellian 

 vacuum. Torricelli corrected also, in 1644, the mistake of Castelli respect- 

 ing the quantities of water discharged by equal orifices, at different 

 distances below the surface of the water in the reservoir. Castelli's ex- 

 periments, made about 1640, were the first of the kind, and some of them 

 really tended to the improvement of the science of hydraulics, but others 

 appeared to show that a double height of the head of water produced a 

 double discharge. Torricelli's more accurate observations proved that a 



* It is an important circumstance in reference to the action of water on lead, that 

 it^is more injurious in proportion to the purity of the water. That which contains 

 less than gg^th of salts in solution, cannot be safely conducted in lead pipes without 

 certain precautions. Christison, Trans. Roy. Soc. Edin. xv. 265. 



f See note, p. 207. t 



