456 Drs. S. M. Copeman and H. W. Hake. [June 19, 
malignant disease, in situations other than the stomach, and, in addition, to 
determine the free hydrochloric acid and some other constituents by methods 
which had also been used by Moore. We therefore commenced in May, 
1907, the chemical examination of a series of test-meals from patients 
suffering from cancer which, by the kindness of Dr. Lazarus-Barlow, of 
Middlesex Hospital, Dr. Murrell, of Westminster Hospital, and Mr. Leaf, of 
the Cancer Hospital, we were able to obtain from time to time. Since the 
above date we have examined 34 test-meals. The results obtained as regards 
physiologically active hydrochloric acid, inasmuch as they appear to confirm 
our conclusions with reference to mice stomachs and to contradict or 
materially modify the conclusions arrived at by Moore and others, will be of 
interest and, in fact, become a necessary adjunct to this research. 
The test-meal given in all cases consisted of one pint of tea and a large 
round of toast. It was given as usual in the morning, fasting, and was with- 
drawn one hour after administration. Ewald, with whom the suggestion of 
test-meals originated, laid down as an essential condition that the test-meal 
should be withdrawn without the addition of water. Apparently, however, 
whether owing to difficulty of manipulation (which in some cases is 
considerable) or from a lack of appreciation of this fundamental rule, 
water is occasionally added by some operators to aid the withdrawal of 
the test-meal. We found that in the test-meals supplied to us, as a matter 
of fact,a number of the earlier ones had been thus diluted, and it was, in fact, 
due to the much larger quantity of hydrochloric acid found in some later 
test-meals (which proved on enquiry to have been undiluted), that our 
attention was especially drawn to this unforeseen contingency. 
Thus in Cases 1 to 4 (Middlesex Hospital), the hydrochloric acid found was 
approximately 0:06 per cent., while in Case 5 (Westminster) we found 
nearly 0°35 per cent. Again, Case 6 (Middlesex) gave us 0:09 per cent., while 
Cases 7 to 9 (Cancer Hospital), gave 0°16, 0°22, and 018 per cent., 
respectively. On enquiry in connection with these greatly differing results 
we ascertained that in Cases 1 to 4 water had been used in order to aid the 
withdrawal of the test-meal. On the other hand we learnt that in Case 5 
and in Cases 7 to 9 the test-meals had been withdrawn without the addition 
of water. 
Owing to repeated assurances that it was not feasible, except in isolated 
cases, to obtain the original undiluted meal, we agreed to the suggestion in 
the first instance with one of the house surgeons at the Cancer Hospital, that 
10 ounces only of water should be used to aid the withdrawal of the test-meal, 
and that if the whole of this amount was not required, the remainder should 
be added to the fluid withdrawn. The withdrawal of test-meals 11 to 23 was 
