4 BULLETIN 1161, U. S. DEPARTMENT OF AGRICULTURE. 
that on the following day or days they would be given their choice 
of the three kinds of ice cream. Sometimes comparison of samples 
was made twice a week, but usually only once a week. The pur- 
chasers, however, could change from one kind of ice cream to another 
at will. 
EFFECT OF FAT CONTENT ON PALATABELITY OF ICE CREAM. 
In this experiment a comparison was made of three ice-cream 
mixes that contained 18, 15, and 12 per cent of fat, while the other 
constituents and methods of manufacture remained the same. 
These three mixes, designated as F, D, and E, respectively, were sold 
on each of four consecutive days for three continuous weeks. 
The total number of sales for the period covered was 551, and the 
number of persons (adults) who purchased was 67. Of this number 
%! ^ * 
1* 
Ajl ffi mm m _ > ...q^H i* * 
SHIP**. i^^^B^.^ 
1 
Fig. 3.— Their second plates in the experiment of "eating at will." Three platefuls were frequently eaten. 
316 sales were made to 27 persons, as shown in Table 1, who purchased 
the ice cream regularly each week during the 3-week period. 
Table 1. — Preference for different proportions of fat in ice cream by 21 purchasers on 
4 days in each of 3 consecutive weeks. 
Mix. 
Times chosen by each purchaser. 
F, 18 per cent fat. 
D, 15 per cent fat. 
E, 12 per cent fat . 
11 11 4 11 11 2 10 10 12 12 10 10 11 7 9 
004014200000022 
004006020022021 
Times chosen by each purchaser. 
Total sales. 
F, 18 per cent fat 
11 8 10 12 9 9 12 6 11 11 7 
04 2 020 05 05 
00 012 00 1 10 
Number. 
259 
33 
24 
Percent. 
82.0 
10.4 
E, 12 per cent fat 
7.6 
Total 
316 
100.0 
