346 Scientific Proceedings, Royal Dublin Society. 



degree. As we know nothing about the causes at work in the observed 

 segregation of radium, it might well be that similar causes operated to segre- 

 gate thorium : such causes might even operate in the latter case with greater, 

 efficiency than in the former. Investigation of the matter is, in short, 

 evidently desirable. 



In the observations which follow I have to acknowledge the very helpful 

 directions of Dr. A. C. O'SuUivan, f.t.c.d., and of Dr. Adrian Stokes, without 

 which I could not have ventured on the investigation. Indeed, I only am 

 to blame if the observations recorded are deemed to be too few in number. 



The procedure adopted was to break up the tissues by maceration in 

 strong HOI over the water-bath. An almost perfect solution was obtained in 

 this way. This solution was then diluted and treated by a method described 

 by me in the Philosophical Magazine for May and July, 1909. This 

 method for measuring the amount of thorium present in a solution is the 

 most sensitive known to me, and is, indeed, the only practical one with which 

 I am acquainted. 



The solution being dealt with is brought to such a dilution as to boil 

 freely. It is enclosed in a flask, and during brisk ebullition a steady current 

 of air is drawn through the flask above tlie surface of the boiling liquid, and 

 thence passes tlirough a condenser, and from this through drying tubes to a 

 gold-leaf electroscope. The condenser removes the steam, etc., returning 

 condensible vapours to the flask ; the air is further dried in the short drying 

 tubes, and finally enters the electroscope with much of the thorium emanation 

 still present. 



If the normal rate of loss of charge by the leaf is accelerated, the signifi- 

 cance of this acceleration is determined by adding to the boiling liquid a 

 small, known amount of a solution of thorium. The acceleration in the rate 

 of collapse of the leaf, produced by this, enables the readings to be calibrated. 



In tliis method the effects due to any radium present may be neglected if 

 the precaution is taken of boiling off the radium emanation before connecting 

 the flask to the condenser. This emanation only very slowly regenerates in 

 the solution, whereas the regeneration of the thorium emanation is very fast. 

 A few recent users of this method substitute mechanical agitation of the 

 liquid — by shaking the flask — for ebullition. For most solutions, however, 

 ebullition offers the most ready and thorough method of agitation. 



The sensitiveness of the results now to be recorded is determined by the 

 following experiment : — • 



Into one of the solutions, which had already been tested, and which 

 possessed a bulk of 900 c.c.'s., a quantity of standard solution of thoriauite 

 was put, containing 16 x 10"^ gram of thorium element. The gain in rate 



