Benzhydrol, an impurity in some samples of cyclizine hydrochloride, interferes with the colorimetric estimation of ergotamine in preparations containing ergotamine tartrate and cyclizine hydrochloride. A method for overcoming this difficulty is described. The inhibition of colour development has been utilised to provide a method for the detection and estimation of benzhydrol in cyclizine hydrochloride.THE use of ergotamine tartrate for the treatment of migraine has resulted in the clinical use of compound tablets containing, besides ergotamine tartrate, a stimulant such as caffeine or a drug to combat sickness or both. In this connection we became interested in the assay of granules, containing ergotamine tartrate and cyclizine hydrochloride. It was found that when powdered granules were shaken with 1 per cent tartaric acid the resulting extract could be assayed colorimetrically for ergotamine using dimethylaminobenzaldehyde solution B.P. : this procedure was adopted for routine work. Some batches of granules, however, gave unexpectedly low results for ergotamine content and we were satisfied that the ergotamine tartrate had been put into the granules and that the assay procedure was faulty. It was discovered that solutions of ergotamine tartrate and cyclizine hydrochloride in 1 per cent tartaric acid gave satisfactory assay results when freshly prepared, but underwent a gradual change when heated on a steam bath and after 30 min. gave no colour with the official reagent. Solutions of ergotamine tartrate alone were unaffected by this treatment. It was decided therefore to avoid heating during the assay of the granules but, despite this, low results for the ergotamine content were still occasionally obtained. We concluded that some samples of cyclizine hydrochloride contained an impurity which interfered with the colour reaction.To test this, a solution of ergotamine tartrate in 1 per cent tartaric acid was divided into eight equal portions. To each was added the appropriate amount of cyclizine hydrochloride to afford a solution similar to that used in the assay of the granules, a different batch of cyclizine hydrochloride being used for each solution. All solutions were then assayed colorimetrically for ergotamine. The results showed clearly that some samples of cyclizine hydrochloride interfered with the colour reaction and some did not.Attempts made to isolate the interfering substance by fractional crystallisation were unsuccessful. When, however, an aqueous solution of "impure" cyclizine hydrochloride was distilled at atmospheric pressure a cloudy distillate was obtained which, after standing for several days, 59 T