4 BULLETIN 612, U. S. DEPARTMENT OF AGRICULTURE. 
Experiments Nos, 518, 514, and 519 to 522 report results obtained 
with the immature dasheen, while the remaining four experiments 
(Nos. 543 to 546) are concerned with the digestibility of the mature 
tubers. Al] experimental conditions were maintained as nearly iden- 
tical as possible in both sets of experiments, and, accordingly, the 
results reported should be directly comparable. 
DETAILS OF THE EXPERIMENTS. 
In general, the methods employed in this investigation were the 
same as those for similar experiments reported in recent publica- 
tions? of this office. Considerable care was exercised in preparing 
and serving the experimental diet so that all the foods eaten by the 
different subjects at the different meals should be of uniform com- 
position. The accessory foods (milk, fruit, butter) were so chosen 
and prepared as to be quite uniform in composition ; they were eaten 
as purchased and were not cooked in any way. 
The dasheen, in sufficient quantities to supply all the subjects for 
the entire period of each experiment, was prepared at the beginning 
of the test period. In every case it was parboiled until the outer por- 
tion could be pierced with a fork, and then baked until the center 
of the tubers was thoroughly softened, a method of preparation 
which experience has shown to be satisfactory as regards quality and 
economy of time and which is in accord with a common practice. 
The entire tuberous structure of the plants was used in every case. 
As was to be expected, it was found that the time required for baking 
the large corms was considerably greater than that necessary for the 
smaller cormels or tubers. When thoroughly baked, the thick skin 
was removed at once and the dasheens were passed through a potato 
ricer while still hot, since they do not “ rice” well when cool. They 
were then thoroughly mixed to insure uniformity and for each test 
a representative sample was reserved for analysis. The 10 experi- 
ments here reported occurred at three intervals—two experiments 
(Nos. 513 and 514), beginning September 28; four experiments (Nos. 
519 to 522), beginning October 10; and four experiments (Nos. 543 
to 546), beginning November 22. The subjects were permitted to 
warm the dasheen before eating it if they so desired, but when they 
did so they were required to leave none uneaten. Butter was eaten 
with the dasheen, each man using, on an average, approximately 110 
grams of butter daily, equivalent to 90 grams or more of butter fat. 
The separation of the feces resulting from the diet was accom- 
plished by means of Jampblack taken in gelatin capsules with the first 
1See list on p. 12. 
