164 H. S. H. WARDLAW. 
ammonia, and 2—4 cc. of magnesia mixture were added 
drop by drop, with constant stirring. After about half an 
hour one-third the volume of the solution of strong ammonia 
was added and the solution allowed to stand over night. 
The precipitate was washed free from chlorides with 1: 3 
ammonia in the usual way and ignited moist as described 
above, to Mg,P.O,, in which form it was weighed. The 
ignited precipitates obtained in this way are oiten not 
white, but grey. The following estimations on a standard 
phosphate solution show, however, that the accuracy of 
the determination is not impaired by this. The standard 
solution of phosphate contained 3°3 gm. of microcosmic salt 
per litre and was standardised by evaporating a known 
volume to dryness, igniting, and weighing the sodium 
metaphosphate formed. 
Volume of| Weight of Weight of Weight of Weight of 
Solution. NaPO, Mea, P,O, POF P,O, in 20 ce. 
40 ce. | 0:0349 gm. ae 0:0484 gm. | 0:02420 gm. 
AQ 4; | O-05D i, po 0:0487 _,, 0:02435 ,, 
2 Le 0:0574 om. | 0:0366 ,, | 0-02436. ,, 
20) 1 55 D205 71s sr. 0:0364 ,, 0:02426 ,, 
1 Ug O°0576 »,, O0367 ss. 0:02440 ,, 
The amounts of P.O; calculated from the weights of 
Mg.P.O, thus agree closely with those obtained from the 
weights of NaPO:. 
The following table gives the percentages of ash, OaO, 
and P.O; found in a series of the deposits obtained by 
spinning milk in a centrifuge. 
It will be seen that the deposits contain approximately 
equal weights of P.O; and of CaO. The average of the 
ratio P,O;/CaO is 1°01. The percentages of these sub- 
stances found in the deposits vary from 3°27 to 4°7Z%. This 
variation is parallel to the variation in the ash content of 
the deposit, the ash itself having a fairly constant compo- 
