216 



added until there was a slight excess. After heating, a boiling 

 solution of ammonia oxalate was added, and the precipitated 

 calcium oxalate allowed to stand overnight. It was then 

 washed carefully with boiling water to remove any excess of 

 ammon. oxalate and received on to a filter paper. The paper 

 was pierced and the precipitate washed into a measuring 

 flask with boiling water and warm dilute hydrochloric acid ; 

 sulphuric acid was then added to dissolve the precipitate 

 completely, and to get the oxalic acid into solution. After 

 making up to 250 cc, the solution was titrated against w/10 

 potassium permanganate, and from this titration the amount of 

 CaCo3 found. 



From several estimations which were made in duplicate 

 it would seem that the method of sampling was not to be 

 trusted. The chitinous covering of the shell, and also its 

 organic matrix, probably did not allow of a homogeneous mixing. 

 It has been thought as well to give the results in Table V, with 

 the results of a second estimation, when these were taken, in 

 brackets. 



The same error of sampling would of course apply to the 

 iron estimations. Here 0-5 grammes of the powder was 

 ignited, dissolved in HCl, and then the iron present was 

 converted into the ferric state by careful addition of potassium 

 permanganate to the solution. 



The estimation was carried out colorimetrically with the 

 aid of potassium thiocyanate, against a standard solution of 

 ferric iron.* 



* §ee Sutton, Volumetric Analysis. 



