Lydia 1M. Dewitt 221
however, were very rare, and by tracing them through the series
of sections, I was generally able to convince myself that they
were merely contiguous with the areas and not continuous.
The appearance of the pancreatic tubules varies so much under
different conditions, and the relation of the tubules to the islands
is, in most animals, so intimate, that it would not be difficult
in almost any section to find structures which might be regarded
as transitional. I have seen none, however, which could not
be explained in some other and more rational way.
As to the size and number of the islands in different stages of
digestion and under different diet conditions, Mankowski says that
the areas increase during digestion, while Hansemann states
that the apparent increase is due only to a more marked differ-
entiation because of the change in the acini themselves.
To test the changes produced in the islands of Langerhans by digestion and diet,
I have examined, measured, and counted a large number of islands .from a
considerable number of guinea-pigs under the following conditions: (I)
normal in full digestion; (2) after different periods of time without food or drink;
(3) after they had been kept on pure carbohydrate diet for different periods;
(4) after the same periods on pure meat diet. As the guinea-pigs had to be
forced to the meat diet, the earlier animals showed purely the changes of in-
anition, and I have not included them in my table. In order to make thc con-
ditions as nearly parallel as possible, I, in all cases except the first, killed thc
animal about fourteen hours after eating, divided the fresh pancreas into three
approximately equal portions, fixed them in saturated bichloride solution; and
cut the sections longitudinally in the direction of the main duct and through
that portion of the gland which included the main duct. The sections therefore
represented approximately the same regions of all the organs examined. I
have measured sections of areas rather than the three dimensions, since the
wvork previously reported has been based on the appearances noted in sections.
The variations in size due to cutting through peripheral portions of the islands
wrere overcome by measuring large numbers and taking the average. I counted
and measured all the islands in the entire section in each case and in at least
five sections, so that my average may certainly represent a true average for the
areas in this region of the pancreas. While some qualitative changes were
noted in the areas,-such as an increase or diminution of the number of
eosinophile cells, a granular change in the cells, atrophy of cells with increase
of intercellular substance,-there were none which could be regarded as constant
for any one experiment and constantly increasing with the duration of the ex-
periment. The table given below gives a brief synopsis of the main data
gained from the counts and measurements. The largest number of very large
islands was seen in the animal which had been kept on a meat diet for sixty
days, while the proportion of island tissue to pancreatic glandular tissue is