IT is often assumed that the placenta acts as an efficient barrier preventing the passage of harmful substances from maternal to foetal blood.To what extent substances, not normally present in the body, yet of a diffusible nature, can be transferred to the offspring via the milk has not been investigated for many substances. If these two points, the transference across the placenta and into the milk are tested by giving the mother some substance capable of taking part in metabolism, the difficulty is to trace its fate. The use of fluorine, it was thought, might get over this difficulty for the following reasons. Fluorine is normally present in animals in very small amounts only, and then chiefly in teeth and bones. It is possible to feed to animals sodium fluoride in suitable doses (0.05 p.c. of the diet) without seriously arresting, though affecting, their growth, development and fertility. Fluorine can be estimated in very small quantities by distillation of the fluorine as H2SiO6, conversion to HF and determination of the latter colorimetrically by its bleaching effect on the red lake of zirconium and alizarin. Since in some cases decomposition products of the perchloric acid, which also bleach the lake, distil over, it is necessary, in order to avoid high values, to do several blank determinations on each sample of the acid.