1, P. Armsby—Digestion Experiments. 365 
Pertop I, 
Vv. 
83°86 per cent. +0°125 00156 
aso © +0°155 0°0240 
83°60 07135 0°0182 
83°59... + —0°145 00210 
34°94 0 0:0789 
Sum, 3 
Average, 82:°735 « 
Probable error of one determination 0°110 per cent. 
Prriop II, 
84°65 per cent. +0°2925 0-085556 
84°52 +0°1625 0°026406 
8t16 > « —0 2275 0051756 
ce 6 Seam —0°2275 0°051756 
0 0°215475 
Sum, 337°4, 
Average, 843575 
Probable error of one determination 0°186 per cent. 
Subsequently, twelve determinations of dry matter were 
made upon another fodder (coarse wheat bran) by the same 
method. 
_ These twelve determinations gave as the probable error of a 
Single determination of dry matter by this method + 0°12 per 
cent, a result agreeing well with the smaller of the two errors 
Computed above, 
f, then, in our computations, we assume + 0°12 per cent to 
'epresent the probable error of one determination of dry matter, 
we shall at least not over-estimate it. ~ é : 
or a computation of the probable error of the nitrogen and 
trude fiber estimations in the fodder we have eight determina- 
hons each, made on samples agreeing so closely in composition 
at they may fairly be me ave been identical, 
On this assumption, we get the following results: 
Nitrogen. Crude fiber. 
210 0 . pee 25 
—0°05 00025 32-92 0-65 0°42 
320 +0°05 00025 32-95 0°62 844 
6d 0-08 0-00 33°55 0-0 0-0004 
209 0-06 00036 33°64 +0°07 0-0049 
221 +006 0-0036 33°71 +014 00196 
neg +0°04 0-0016 33°44 —0'13 00169 
221 +0°06 0-0036 33-96 +0.39 0°1521 
213 —0-02 0-0004 34°36 40°79 0°6241 
a ere LED ae FROGS TA Pee 
c 0 0-0242 268-56 0 16249 
Probable error of 1 determination Probable error of 1 determination 
+0-040, + 0°325, 
Twelve determinations each of nitrogen and crude fiber on 
the Sample of bran mentioned above gave the following results: 
AM. JouR. Sct.—Turrp Senes, Vou. XXIX, No. 173, May, 1885. 
as 
