corros. sublimate 2,5 gm., picric acid 1 gm., 15 cc. 40°, formalin) 
for at least twelve hours, but in some cases duplicates were fixed in 
ZENKER’S fluid and in one case, aceto-bichromate (glacial acetic 5 cc., 
3°/, solution of potassium bichromate in distilled 95 cc.). In all cases 
the glands were then stained in bulk either in EnrLicH’s haematoxylin 
or GRENACHER’S borax carmine, embedded, sectionized, and the sections 
differentiated and stained on the slide for a few minutes with a solution 
of picro-indigo-carmine (see formula in 15), diluted with an equal volume 
of 70%, alcohol. These methods of fixation and subsequent staining 
give most excellent results, the color contrasts of the different tissues 
often being very striking. In none of these slides was I able to detect 
haemolysis. 
I determined to make the following series of blood smears in 
order to test the evidence afforded by the sections. All those fish 
which were weighted had been so for six hours. 
1) Smears made of blood (arterial and venous) from activated 
gland. 
2) Smears made of blood (arterial and venous) from caudal artery 
and vein of same fish (control smear). 
(Three fish treated this way.) 
3) Smears made of blood (arterial and venous) from the inactive 
gland of one non-weighted fish (control smear). 
4) Smears made of blood (arterial and venous) from caudal artery 
and vein of same fish (control smear). 
I made several smears from each source of blood and fixed some 
(for 5 minutes or so) with 1°, osmic acid vapour, subsequently staining 
with EHRLICH, and others (for 4 minutes) in a fixative recommended 
I believe by Prof. Mincutn, viz. 100 cc. absolute alcohol, 1 gm. corros. 
sublimate, 0,5 cc. glacial acetic, and stained subsequently with EHRLICH 
and picro-indigo-carmine and in some cases with the ISRAEL-PAPPEN- 
HEIM stain to be referred to below. With both fixatives I obtained 
good preparations. In no case could I detect haemolysis. 
In two very large Pollack (weighing roughly I should say be- 
tween 500 and 600 grams) I managed to obtain smears of the blood 
taken solely from the vein connected with the active gas gland, but 
in these also I failed to detect haemolysis. 
Again it occurred to me that if instead of making smears of blood 
which is circulating through the gland capillaries and is therefore ex- 
posed to the action of the hypothetical lysin secreted by the gland 
for only a short time, I ligatured the gland artery and vein and so 
enabled this lysin, if it exists, to have more effect on the stationary 
