350 BUTTER. 



difierenoes of temperature, the author finds that the readings 

 are not strictly correct. 



A chart for the correction of butyro-refractometer readings 

 for temperature may be constructed thus : — Select a sheet of 

 squared paper at least 120 units by 200 units wide ; at a point 

 34 units from the bottom, set out horizontally a series of points 

 5 units apart, and at a point 119 units from the bottom a similar 

 series of points 7 units apart ; join the corresponding pairs of 

 points to form a series of temperature lines. The middle vertical 

 line is selected as the standard temperature, say 35°, and each 

 line to the right will represent a temperature 1° higher, and 

 to the left 1° lower than the next preceding line. 



From the bottom at a point 100 units from the standard 

 temperature line to the left draw a line to the point which lies 

 20 units from the bottom and 100 units to the right of the standard 

 temperature line ; this will represent 0° on the scale of the 

 butyrometer ; draw parallel to this a series of lines 10 units 

 apart measured vertically, and mark these 10°, 20°, etc., of the 

 refractometer scale. To use the chart, find the point of inter- 

 section of the lines corresponding to the observed temperature, 

 and scale lines (differences between the lines can be estimated 

 with sufiicient accuracy by the eye), and the distance measured 

 horizontally between this point and the vertical standard line 

 will give the correction to be added if on the right, or subtracted 

 if on the left ; 10 units of distance equal one scale degree of 

 correction. 



This chart is easy to make, and still easier to use, and the 

 author has found it to give very accurate results over a con 

 siderable range of temperature, not only for butter, but also 

 for other oils and fats, and for the standard fluid. 



The author has found that genuine butters vary from 43 '7° 

 (in a sample giving a Reichert value of 16 '0 c.c.) to 49 "0° (in a 

 sample giving a Reichert value of 10"5 c.c), and average 46"0° 

 at a temperature of 35° C. The value 47 '0° has been proposed 

 as a practical limit. 



The equivalent at other temperatures of this limit is as follows: — 



Some importance has been attached to the colour observed at 

 the edge of the dividing line, and a blue colour has been alleged 

 to be indicative of margarine. In the author's experience this 

 property is valueless. Thus the sample giving a reading of 43 "7° 

 was tinged red, and that giving 49 "0° was tinged blue, though 

 both were authenticated as genuine butters. 



