TRANSACTIONS OF SECTION I. 895 
is mentioned, all that is meant by the expression is that the blood or urine respec- 
tively contains so much reducing material, estimated as glucose. My method in 
extracting the sugar from the blood has consisted in boiling with sodium sulphate, 
as described by Pavy in his book on ‘ The Physiology of the Carbohydrates.” The 
reducing power of the blood and urine I have estimated by the ammoniated 
Fehling’s solution recommended by Pavy, and described in the same book. I 
have convinced myself of its accuracy, as in the estimation of standard solutions 
of sugar I have always obtained a value correct to the first two places of 
decimals. 
I have obtained the following results :— 
1, That if the thoracic lymph from a dog which has been starved a day is 
injected into the portal system of a cat, no glycosuria results, nor is there any 
hyperglyceemia beyond the degree which sometimes results from the use of 
anesthetics, that is from about *2 per cent. to -28 per cent. 
2. That if the thoracic lymph from a digesting dog is injected into the portal 
system of a cat, glycosuria of a degree varying from 1 per cent. to 9 per cent. and 
hyperglycemia of a degree varying from ‘3 per cent. to ‘9 per cent. result. 
3. That if thoracic lymph from a digesting dog be injected into the systemic 
circulation of a cat, hyperglycemia and glycosuria again result, though not in as 
high a degree as after portal injections; that is, hyperglycemia of a degree from 
‘3 per cent. to ‘5 per cent., and glycosuria of a degree from 1 to 3:75 per cent. 
4, That if the thoracic lymph from a digesting cat be injected into its own 
splenic vein, hyperglyczemia and glycosuria result. 
’ 5, That the mere injection of a few bubbles of air with a little normal salt 
solution into the portal system of a cat is followed by glycosuria; while the 
injection of salt solution or of serum from dog’s blood is followed by no glyco- 
ps and by no hyperglyczemia, beyond the slight degree due to the anesthetics 
employed. 
%G That, as Biedl proved, the mere formation of a fistula of the thoracic duct 
in a dog is followed by glycosuria ; and whereas he only experimented on the dog, 
and usually obtained glycosuria of a degree not above 2 per cent. and never 
more than 5'8 per cent , I have also obtained this result in the cat; and in a dog 
which had been starved thirty-six hours I obtained glycosuria of a degree of 
10°52 per cent. 
My explanation of these various results is that the internal secretion of the 
pancreas’ passes into the blood chiefly wd the thoracic lymph, while some toxie 
substance, formed probably in the intestine, is present in the thoracic lymph 
during digestion, but is absent therefrom during starvation, though it probably is 
present also in the portal blood both during digestion and starvation. This toxic 
body and the internal secretion of the pancreas are antagonistic to one another 
In the absence of the internal secretion of the pancreas, this substance has a toxic 
action on the glycogenic tissues of the body whereby they cannot, to the same 
extent as normal, form glycogen out of sugar, and probably are stimulated to 
convert glycogen into sugar. 
Thus the injection of thoracic lymph from a digesting dog causes glycosuria 
because it contains the toxic substance. On the other hand, leading the thoracic 
lymph away from a fasting dog is followed by glycosuria, because, the internal 
secretion of the pancreas being thus removed, the toxic substance in the blood is 
given free play. 
I attribute the glycosuria following the mere injection of a few bubbles of air 
into the portal circulation of a cat to the fact that this air very largely accumu- 
lates in the veins of the pancreas; as a result of which the pancreas is probably 
crippled in forming its internal secretion. 
Though the above explanation is nothing more than a supposition, yet it has 
very great probability in its favour, because it harmonises the glycosuria caused 
by interference with the thoracic duct with the glycosuria following excision of 
the pancreas; and thus gets rid of the necessity of our calling the former ‘a new 
form of experimental diabetes,’ as was done by Bied]. 
