181 3. j Definite Proportions in Chemical Combinations. 301 



manganese, and platinum : for we find the deutoxide of these 

 metals combine not with 1 but with 2 integrant particles of 

 sulphuric acid. Hence it is probable that the sulphate of iron 

 given in the table does not exist. 



If we examine the notes under the table of the composition 

 of the salts, it will be seen that none of the other six exceptions 

 to the canon of Berzelius can be considered as valid; because 

 thev are either stated from mere theory, or they do not agree 

 with the analyses of the salts in question hitherto published. 



Upon the whole, then, the sulphates furnish one decisive 

 exception to Berzelius's canon. This canon appears to me at 

 present entirely empyrical. I cannot see any sound reason 

 (different from the result of analyses) that should lead us to 

 adopt it. Hence the knowledge of this exception disposes me 

 at present to reject it ; but we shall be able to judge with more 

 correctness after we have examined a few more genera of salts, 

 especially the nitrates, upon the analysis of which Berzelius 

 seems to me to have founded it. 



(To be continued.) 



Article XII. 



Magnelical Observations at Hackney Wick. By Col. Beaufov. 



Magnet ical Observations. 



Latitude 51" 32' 40" North. Longitude West in Time 6"-^*^- 



1813. 



Mean of ("Morning at 8>> 44' Variation 84° 15' 58" 



ObfcTvaiium 1 N .on at 2 02 Ditto 84 23 32 



in Aug. (.livening at 7 05 Ditto 24 16 OH 



