58 ANNUAL OF SCIENTIFIC DISCOVERY. 



This was selected as the strongest form of boiler in the time of Xcwcomen 

 and Leigh ton; and it was, probably, for a similar reason that the glass- 

 blower forms the bottom of bottles with an elevated cone penetrating for 

 some distance into the interior of the cylindrical part. This gives great 

 strength to the bottle in resisting internal pressure, and at the same time 

 reduces the quantity of liquid contained in the bottle; a consideration inde- 

 pendent of strength, and probably a matter of no small importance to the 

 retail dealers in wine and ardent spirits. 



The result of these experiments upon metal tubes subsequently suggested 

 to Mr. Fairbairn the propriety of similarly testing the resisting powers of a 

 perfectly homogeneous, crystalline, and rigid material; in order that our 

 knowledge of the laws that govern the resistance of vessels to collapse might 

 be confirmed and extended. Glass was selected, not only because of its ful- 

 filling better almost than any other material the conditions sought for, and 

 from the ease with which it could be manufactured into the forms required, 

 but also because it was hoped that the results would be of practical value in 

 those cases in the arts and in experimental science, in which it is so exten- 

 sively employed. The experiments were conducted in a similar manner to 

 those upon iron. Some cylinders and globes, blown out of good flint-glass, 

 were procured direct from the maker. The open ends were hermetically 

 sealed by means of the blowpipe; and the globes, etc., were placed in a strong 

 wrought-iron vessel capable of sustaining a pressure of 2,000 Ibs. to the square 

 inch. Water was pumped in by means of a force-pump; the pressure was 

 recorded by a Schaffer's gauge; and the point of rupture was indicated by an 

 explosion within the vessel, and by a sudden decrease of pressure. The first 

 experiments were upon glass globes, intended to be perfectly spherical, but 

 in most instances somewhat flattened upon the side opposite to that from 

 which they were blown. Notwithstanding this ellipticity, some of the globes 

 bore enormously high pressures, especially when the extreme tenuity of the 

 glass was considered, amounting, as it did, from only _1 ._ to y^ of an inch 



in thickness. In one instance seven globes of glass were submitted to test, 

 three of which were intended to be 5 in. diameter, one 5 1-2 in. and three 8 in., 

 but varying as before mentioned. The bursting pressure of the first four was 

 292 Ibs., 410 Ibs., 470 Ibs. and 475 Ibs. to the square inch, equivalent in the last 

 case to twenty tons upon a 5 1-2 in. globe, 1 in. thick, before it was frac- 

 tured. The 8 in. globes burst respectively at 35 Ibs., 42 Ibs. and CO Ibs. to an 

 inch ; but they were unfortunately all elliptical to a serious extent, the diam- 

 eters of that which burst at 42 Ibs. being respectively 8'20 in. and 7'30 in. 



In experimenting with homogeneous glass cylinders, blown with hemis- 

 pherical ends, it was found that the law deduced from the experiments upon 

 iron tubes, applied equally well to the glass, viz., that the strength of cylin- 

 drical vessels, exposed to a uniform external pressure, varies inversely as to 

 length. Thus a glass cylinder 4'06 in, diameter, 13 3-4 in. long and '045 in. 

 thick, collapsed under a pressure of 180 Ibs. to the square inch ; while another, 

 4'05 in. diameter, 7 in. long, and '040 in. thick, yielded only under a pres- 

 sure of 380 Ibs. to the inch. 



EVAPORATIVE POWER OF BRASS, COPPER, AND IRON BOILER-TUBES. 



A late number of the London Mechanics' Magazine contains an article on 

 the above important question, by W. G. Tosh, from a paper read by him 



