110 Prof. F. W. Clarke on the Results obtained 

 Table of Atomic Weights {continued). 





H=l. 



= 16. 





H=l, 



0=16. 



Oxygen 



15-9633 



16-0000 



Tantalum ... 



182-144 



182-562 



Sulphur 



31-984 



32-074 



Scandium ... 



43-980 



44-081 



Selenium . . . 



78-797 



78-978 



Yttrium 



89-816 



90-023 



Tellurium . . . 



127-960 



128-254 



Erbium 



165-891 



166-273 



Chromium . . 



52009 



52-129 



Ytterbium . . . 



172-761 



173-158 



Molybdenum 



95-527 



95-747 



Cerium 



140-424 



140-747 



Tungsten ... 



183-610 



184-032 



Lanthanum 



138-526 



138-844 



Uranium ... 



238-482 



239-030 



Didymium 



144-573 



144-906 



Manganese . . 



53-906 



54-029 



Carbon 



11-9736 



120011 



Iron 



55913 

 57-928 



56-042 

 58-062 



Silicon 



Titanium ... 



28-195 

 49-846 



28-260 

 49-961 



Nickel 



Cobalt ..... 



58-887 



59-023 



Zirconium . . . 



89-367 



89-573 



Copper 



63-173 



63-318 



Tin 



117-698 



117-968 



Boron 



10-941 



10-966 



Thorium . . . 



233-414 



233-951 



Aluminum .. 



27-009 



27-075 



Platinum ... 



194-415 



194-867 



Gallium 



68-854 



68-963 



Iridium 



192-651 



193-094 



Indium 



113-398 



113-659 



Osmium 



198-494 



198 951 



Nitrogen . . . 



14-021 



14-029 



Palladium... 



105-737 



105-981 



Antimony ... 



119-955 



120-231 



Rhodium ... 



104-055 



104-285 



Bismuth ... 



207-523 



208-001 



Ruthenium 



104-217 



104-457 



Columbium 



about 94 



about 94 



Gold 



196-155 



196-606 





Here we have sixty-six elements, or, rejecting columbium 

 as too vaguely determined, sixty-five. Such elements as 

 phillipium, decipium, thulium, samarium, &c. are not yet 

 sufficiently well known to be considered in this connexion. 



In his recent superb investigation of the atomic weight of 

 aluminum, Mallet makes substantially the following argument 

 in favour of Prout's hypothesis. Considering the atomic 

 weights of eighteen elements only as well determined, he finds 

 that ten of them have values varying less that 0*1 from whole 

 numbers. In other words, these ten elements have atomic 

 weights varying from even multiples of that of hydrogen by 

 insignificant amounts. What is the probability that this 

 agreement with Prout's hypothesis in ten cases out of eighteen 

 is purely accidental, as those hold who agree with the views 

 of Stas ? Working this problem out, he finds the probability 

 of mere coincidence to be 1 : 1097*8 ; and he concludes that 

 Prout's hypothesis is still worthy of careful consideration. 



Applying Mallet's reasoning to the Table of atomic weights 

 now before us, we find that in the first column, when H = 1, 

 twenty-five out of sixty-five elements have atomic weights 

 falling within one tenth of a unit of whole numbers ; but 

 many of the figures which fall outside this limit of variation 

 involve the variation of oxygen multiplied many times over. 

 We must therefore study the second column, which assumes 

 = 16. Here we have thirty-nine elements falling within 

 the limit of variation assigned byMallet and twenty-six falling 



