21

It has been shown ⁽¹⁾ that in pathological conditions the salinity,
        measured in this way, is greatly increased in œdema and anæmia.
        The lower percentage of hæmoglobin present in the blood of the
        Bengali would appear to have a distinct connection with the higher
        salt concentration of the serum; the salinity of the serum varying
        inversely as the hæmoglobin value of the blood. What the ex-
        planation of this is we are unable to say. It would appear, how-
        ever, that in those diseases in which the salt concentration of the
        serum is markedly increased —anæmia, Bright's disease, etc.—the
        red blood corpuscles are more resistant to a dilution of the serum
        than in health. That this is not due to a retention of chlorides
        by the blood is evident from the fact that, although estimation of
        the total chlorides of the serum by the Harvey-McKendrick
        method shows a decided increase, the increase is nothing like
        sufficient to account for the very high percentage obtained by
        Wright's hæmolysis method. Some other factor, at present
        undetermined, increasing the resisting power of the red blood
        corpuscles and which cannot be explained by osmosis or mere isoto-
        nicity, must be present. However, in health or even in the degree
        of anæmia met with in the Bengali, the results can be completely
        explained by osmosis; the proof of this statement will be found
        in the close agreement of the percentage of total salts estimated by
        the ordinary quantitative means and by Wright's hæmolysis
        method.

iii. The chlorides of the serum—

Voila's average value for human blood serum—0.55 per cent, NaCl—agrees
        very closely with the faultless results obtained by Schmidt,
        Wannach and Biernacki.⁽²⁾ This is the quantity usually accepted
        for the blood serum of Europeans. Landois ⁽³⁾ gives the total in-
        organic salts of human blood serum at .85 per cent., the chlorides
        alone being .50 per cent.

As we have already stated, the average percentage of chlorides found
        in the blood serum of the native of Bengal varies within narrow
        limits in health and is usually from 0.72 per cent. to 0.75 per cent.
        We tabulate a few of the estimations carried out showing the total
        chlorides of the serum of the Bengali.

                ⁽¹⁾ M'Cay—The Lancet, 1st June 1907.

                ⁽²⁾ Von Noorden—The Physiology of Metabolism, 1907.

                ⁽³⁾ Landois—A text-book of Human Physiology, 1904.