A FEW months ago when working up slops of salt-free iso-electric ovalbumin solutions from half a year of experiments, we made the surprising observation that the protein, when crystallized with ammonium sulphate, appeared as beautiful rectangular plates instead of the usual flat cigar-shaped needles. We did not at that time observe any infection of the solutions either by bacteria or moulds ; but since urea was present and no antiseptics had been added, infection was a factor to be reckoned witheven at the low temperature (1°) at which the protein had been stored. A further perusal of our stock of old dialysed ovalbumin solutions in the refrigerator revealed six samples, one to two years old, three of which had been stored with toluene, three without. Only the latter were infected, two with bacteria and one with moulds (and bacteria ?), and upon crystallization they gave a good(60 per cent) yield of plates. All attempts to isolate plates from the former solutions were in vain.