140 ANNUAL REPORT. 
acid that is in the solution, and furnish slight excess, some additional ammonium chlor- 
ide is also added at this point. ‘The beaker is set on the hot plate and the solution boiled 
a moment until it has ceased foaming. Without waiting for it to settle it is filtered by 
suction into an 800 cc. Griffin beaker through an 11 cm. ashless filter. After rinsing 
out the beaker the filter is washed three times with hot water. The filter is burned 
and the iron oxide and alumina weighed. 
The solution in its larger beaker and about 30 ce. of a saturated solution of ammo- 
nium oxalate in a smaller one are then set on the hot plate and after both are boiling the 
latter is added to the former, which is boiled a moment longer. Jt is immediately 
filtered with the filter pump and bell jar through an 11 em. filter into a little flask, the 
beaker well rinsed out and the filter washed twice, then transferred with the funnel to 
a filtering flask and the washing continued. The filter is then removed from the fun- 
nel, opened out flat on the palm of the hand and the oxalate of lime washed back into 
the beaker in which it was precipitated. About 30 cc. 1 to 3 sulphuric acid and 300 cc. 
hot water are added to dissolve the oxalate, and the lime is determined by titration 
with standard permanganate. ‘This solution is of such strength that using 0.5 g. of 
sample each cc. is equivalent to one per cent. of lime. 
To the filtrate in the flask, cooled in running water, is added about 10 ce. of a strong 
solution of microcosmic salt and 100 ce. of strong ammonia, a rubber stopper inserted 
and the whole well shaken for five minutes. This throws down the magnesium am- 
monium phosphate quite dense and granular, with no tendency to stick to the flask. 
After standing half an hour the solution will have separated out clear about half an 
inch. It is then filtered with the filter pump, rinsing out the flask and washing well 
with 1 to 6 ammonia, the filter is wet with a strong solution of ammonium nitrate 
ignited, weighed and the magnesia calculated as usual. 
INSOLUBLE RESIDUE IN RAW CEMENT MIXTURES AND BURNT CEMENT. 
In raw cement mixtures it is important to estimate the so-called 
insoluble residue for the purpose of determining whether or not the 
erinding has been carried sufficiently far so as to bring about complete 
or sufficient chemical union of the cement ingredients. It is of special 
value in new plants, particularly those working with a limestone-clay 
mixture. It is not necessary to carry on this determination after the 
rate of feeding the tube mills has been adjusted, but it will always prove 
a valuable check on the work done by the grinding machines. 
The analysis proper simply consists in heating a one-gram sample 
over the blast lamp for 20 minutes, treating it with hydrochloric acid 
and alkaline solutions, as described repeatedly under the head of pozzu- 
olanes and rational clay analysis. Properly ground raw mixtures should 
show little or no residue. Any appreciable residue is due to insufficient 
erinding. Of course a sieve analysis would show the condition of 
fineness of a mixture, but, after all, could not differentiate the real condi- 
tion governing chemical union as well as does this mineral analysis. 
Analogous to the determination of the “insoluble residue” in the raw mix- 
ture, we can apply the same method to the burnt cement, with the excep- 
tion that here the carbon due to unconsumed coal mingled with the 
cement must be burnt out. If this is done either before the analysis 
or at the end, before weighing up the residue, the latter will be found 
to consist of any coarse quartz not united with the lime, and any ash 
