ROUTINE ESTIMATION. 103 



water, and is usually pure enough for all purposes. It is liable, 

 however, to contain traces of copper, if this metal is employed 

 in the construction of the coils, condenser, etc. ; if the boilers 

 prime to any extent, it will also contain impurities (salts, scale- 

 preventing composition, etc.), derived from the boiler. 



The author was informed by Hehner many years ago that 

 he had found the viscosity of milk to be proportional to the total 

 solids, and recently Taylor has confirmed this. 



Routine Estimation of Total Solids, A number of dishes 

 or capsules, each marked with a separate number and weighed 

 previously, and a pipette marked to discharge 5 grammes 

 of milk of a specific gravity 1-032 at a temperature of 60 F., 

 are necessary. The dishes should be, if possible, of platinum, 

 and at least 1J inches internal diameter, flat bottomed, and 

 with a rim about J of an inch wide ; this rim should be 

 considerably wider at one side so as to form a lip on which 

 the number should be legibly stamped. These dishes weigh 

 about 12 grammes and cost about 10 each. They may, how- 

 ever, be replaced by porcelain dishes of about 2J inches diameter, 

 glazed all over ; these may be marked by scratching the number 

 on the side with a new file, painting this over with a solution 

 of platinum chloride, wiping oft' the excess, and igniting the 

 capsule ; the number will be marked in platinum. 



The dishes must be previously weighed on a balance accurate 

 to 1 milligramme, and the weights recorded on a table which 

 should be kept in the balance case. It has been found in practice 

 that, if the dishes be carefully cleaned, monthly weighings of 

 the dishes are sufficient, the average loss in this period for dishes 

 used daily having been found to be about 1 to 2 milligrammes. 



The dishes should be arranged in a shallow tray a photo- 

 graphic dish is suitable according to their numbers from left 

 to right, six or seven in a row, beginning at the bottom i.e., 

 the side nearest the operator. This arrangement is chosen so 

 that any milk accidentally dropped from the pipette will not 

 fall into a dish which has been filled previously, but into an 

 empty dish which can be wiped easily. The tray containing the 

 dishes should be on the left of the tray in which the samples are 

 placed. The samples should be arranged in order in rows, 

 beginning at the left-hand bottom corner and going upwards 

 i.e., away from the operator and should, if in cans, have their 

 lids turned back over the next sample can. The taking of 

 samples for the estimation of total solids is much facilitated by 

 having an assistant to stir the milk. The measurement of the 

 quantities of milk for analysis is done as follows : The assistant 

 stirs No. 1 can, and when the cream has been mixed the chemist 

 plunges the pipette into the milk and sucks it up till it enters 



