THE preparation of crystalline secretin has already been described [Hammarsten et al. 1928;Agren & Wilander, 1932; Hammarsten et al. 1933a, b] and a full description of its properties has been given [Agren, 1934, 1935 a, b, Agren et al. 1936 a, b, c, 1937a. The secretin used for the experiments described in the present paperwas the chloride or phosphateprepared from crystalline secretin. These salts were amorphous.An analytical digestion of secretin with aminopolypeptidase and with carboxypolypeptidase seemed to us to be of special interest on account of the ninhydrin reaction with secretin before and after hydrolysis, its isoelectric point, and the fact that one amino-group can be split off by van Slyke determination without loss of activity. Secretin is a basic polypeptide, which does not give the ninhydrin reaction. It can be hydrolysed with pepsin or trypsin, and at higher temperatures in alkaline or acid solution. In consequence it is inactivated and now gives a strong ninhydrin reaction. As far as we know, this combination is unique, and we have compared the ninhydrin reaction of amino-acids with that of secretin in a series of concentrations, secretin not giving the slightest trace of colour, even in very high concentrations and after evaporation to dryness. The reaction is very sensitive to the hydrogen-ion concentration and pH 6-8 is the optimum for development of colour. This was therefore used. After, the determination of the isoelectric point, it was clear that secretin was an amphoteric substance and by means of digestion with carboxypolypeptidase we hoped to determine the nature of the acid group. Aminopolypeptidase was prepared according to Agren [1937]. The preparation was made from hog's gastric mucosa in order that a trypsin-free glycerine solution might be obtained. The