flo- Scientific Proceedings, Royal Dublin Society. 



decigramme divided by a unit-XXIV, i.e. & XXV th gramme. The 

 chemical atom is half of this. Hence the mass of a chemical atom 

 of hydrogen may be taken to be somewhere about half a twenty- 

 fifth-gramme. There is no advantage in retaining the coefficient 

 half in an estimate in which we are not even sure that we have 

 assigned the correct power of 10, and I will therefore, for the 

 sake of simplicity, take the XXVth gramme as being such an 

 approach as we can attempt to the value of the mass of an atom 

 of hydrogen. 



Now io has been ascertained by experiment that for every 

 Ampere of electricity that passes ninety-two sixthgrammes, i.e.. 

 ninety- two millionths of a gramme of water, are decomposed (see 

 B.A. Report 1863, p. 160). This water is the result of a secondary 

 action in the voltameter, but that does not eifect the present 

 inquiry. Ninety-two Ylth grammes of water contain about one 

 Yth gramme of hydrogen, which is therefore the quantity evolved. 

 The metric unit of electricity e^ is 100 Amperes, and will there- 

 fore set free 100 Vth grammes of hydrogen, i.e., one milligramme. 

 Now it appears, from the last paragraph, that this quantity of 



XXV 



hydrogen contains ^^^^^ atoms, i.e., XXII. atoms. And as there is 



a bond ruptured for each atom of hydrogen set free, this is also 

 the number of bonds broken. In other words the quantity of 

 electricity corresponding to each chemical bond separated is 



^' =xxn'' ^^)- 



Collecting our numerical results they are 



Yj = 3. VIII, metres per second... .(1). 



■^' ==ixTiT' <^^>' 



^' = xii' ^^)- 



= ^^^ Amperes (4). 



We have thus obtained approximate values in known measures 

 for the three great fundamental units offered to us by Nature, 

 upon Avhich may be built an entire series of physical units 

 deserving of the title of a truly Natural Series of Physical Units. 



It now only remains to deduce the units of length, time, and 

 mass belonging to this series. For this purpose we may use 



