No. 2.] BRITISH ASSOCIATION MEETING. 161 



relations which subsist between the chemical composition and re- 

 fractive power of bodies for light. He then proceeded — A happy 

 modification of the ice calorimeter has been made by Bunsen. 

 The principle of the method — to use as a measure of heat the 

 change of volume which ice undergoes in melting — had already 

 occurred to Herschel, and, as it now appears, still earlier to Her- 

 mann ; but their observations had been entirely overlooked by 

 physicists, and had led to no practical results. Bunsen has, in- 

 deed, clearly pointed out that the success of the method depends 

 upon an important condition, which is entirely his own. The ice 

 to be melted must be prepared with water free from air, and must 

 surround the source of heat in the form of a solid cylinder frozen 

 artificially in situ. Those who have worked on the subject of 

 heat know how difl&cult it is to measure absolute quantities with 

 certainty, even where relative results of great accuracy may be ob- 

 tained. The ice calorimeter of Bunsen will therefore be welcomed 

 as an important addition to our means of research. Roscoe has 

 prosecuted the photo-chemical investigations which Bunsen and he 

 began some years ago. For altitudes above 10 degrees, the rela- 

 tion between the sun's altitude and the chemical intensity of lio-ht 

 is represented by a straight line. Till the sun has reached an al- 

 titude of about 20 degrees, the chemical action produced by difi"u- 

 sed daylight exceeds that of the direct sunlight ; the two actions 

 are then balanced, and at higher elevations the direct sunlight is 

 superior to the diffused light. The supposed inferiority of the 

 chemical action of light under a tropical sun to its action in higher 

 latitudes proves to be a mistake. According to Boscoe and Thorpe, 

 the chemical intensity of light at Para, under the equator, in the 

 month of April, is more than three times greater than at Kew in 

 the month of August. Hunter has given a great extension to the 

 earlier experiments of Saussure on the absorptive power of char- 

 coal for gases. Cocoanut charcoal, according to Hunter's experi- 

 ments, exceeds all other varieties of wood charcoal in absorptive 

 power, taking up at ordinary pressures 170 volumes of ammonia 

 and 69 of carbonic acid. Methylic alcohol is more largely ab- 

 sorbed than any other vapour at temperatures from 90*^ to 127^^, 

 but at 159^^ the absorption of ordinary alcohol exceeds it. Cocoa- 

 nut charcoal absorbs 44 times its volume of the vapour of water at 

 127^. The absorptive power is increased by pressure. Last year 

 two new processes for improving the manufacture of chlorine at- 

 tracted the attention of the section ; one of them has already proved 

 Vol. VI. c No. 2. 



