124 MR JOHN SCOTT ON THE BURNING MIRRORS OF ARCHIMEDES. 



more than 100 feet, produced a heat which could scarcely be endured. Apparently 

 convinced of the practicability of the achievement by means of plane mirrors, he 

 entreats future mathematicians to prosecute the subject. Buffon, following in his 

 steps, completely established the fact that combustible materials can be set on 

 fire at distances corresponding to the accounts we have of the mirror of Archi- 

 medes. This he effected by means of a combination of plane reflectors, consist- 

 ing of ordinary looking-glasses, 8 inches by 6, attached to a single frame, each 

 glass, as well as the supporting frame, being capable of motion in every direction. 

 With forty of these glasses he set on fire tarred beech at a distance of 66 feet. A 

 plank, smeared with tar and brimstone, was ignited at 126 feet by 98 glasses. A 

 combination of 128, with a clear sun, inflamed very suddenly a plank of tarred 

 fir at 150 feet, the conflagration springing up at once over a space of 1G inches 

 in diameter — the whole reflected image of the sun at that distance. In addition 

 to these experiments made about the beginning of April, others were exhibited 

 with the summer sun, by which wood was kindled at 200 and 210 feet, and 

 silver and other metals were melted at distances varying from 25 to 40 feet. 



Let us now consider the evidence on the opposite side of the question. Des- 

 cartes and others have treated the whole affair as fabulous, from the belief that 

 the burning glass must have consisted of a single spherical or parabolic reflector. 

 But since no mention is made of the kind of specula employed by Archimedes, 

 such objections, after the successful experiments of Buffon, necessarily become 

 irrelevant. Another and quite different ground of doubt has arisen from the circum- 

 stance, that Polybius, Livy, and Plutarch make no mention of the destruction 

 of the Roman fleet by means of burning glasses, although they describe somewhat 

 in detail the ballistse and other military engines constructed by Archimedes to 

 resist the assailants. How far the silence of the forementioned writers should be 

 taken as negative evidence, it is not easy to determine. It seems to indicate, at 

 least, that the damage effected by the burning mirror was confined to compara- 

 tively narrow limits ; for, if prominent in the defence, we naturally expect that 

 it would have received special notice. But allowing that the fleet sustained no 

 serious injury from this source, some fact must be left sufficient to account for 

 the belief prevalent among ancient authors in favour of the achievement. 

 Nothing less, I conceive, will suffice than to admit that Archimedes made an 

 attempt to destroy the Roman fleet in the manner described. Such an admission, 

 however, implies that he must have previously tried his combination of reflectors 

 in private, and was able to ignite combustible substances at considerable dis- 

 tances. The mere attempt to bring an artillery so singular and subtle to bear on 

 the fleet is in itself a conclusive proof that an experiment, similar to those 

 exhibited by Buffon in the Jardin-du-Roi at Paris, had been successfully per- 

 formed 2000 years before within the walls of Syracuse. We can no more suppose 

 the contrary than believe that any of the nations of modern Europe would send 



