With Plate 6 and 6 Text-figures)The literature relating to the growth of moulds at different humidities is surveyed.Observations were made on the development of mould growh on six widely different types of feeding-stuffs for storage periods extending over some 31 years. The main factors controlling mould growth were found to be : I . Thc relative humidity rather than the moisture content of the food. 2. The length of the rtoruge peziod. Mould growth took place relatively quickly on all feedhg-stuffs stored at 100-75 % R.H. Below 75 R.H. mould growth, while not prevented, only developed after a very prolonged latent period. Mould development was observed on locust beans exposed to a humidity as low as 65 "/o after a latent period of over z years.3. The b a k e and type of nutrients provided by the various feeding-stuffs mas shown to influence the latent period as well as the extent of mould deterioration. This was confirmed by experiments on artificially dried grass fractions and by experiments on starch, protein and fibre mixtures.4. The tempnaiure of rtoruge was found to affect the rate of mould development; mould growth developed sooner on samples stored a t zz°C. than at 15.5'C.5. The type o j mould rpecier present. At high humidities all mould species grew rapidly: at low humidities, only a few types were capable of growth.

The water uptake of a wide variety of feeding-stuffs was determined over a range of fixed humidities. It was shown that the level and shape of the water uptake curves are closely related to the amounts of soluble carbohydrate and protein present. Fibre exerts a depressing effect on the water uptake, while the value is also lowered by the presence of inert fats and non-hygroscopic ash constituents. These conclusions were conhned by direct experiments on the water uptake of samples of pure starch, protein and fibre.The results of the investigations have enabled safe limits of moisture content to be laid down, for both short-and long-period storage, below which mould growth w i l l not normally take place. These limits vary for each individual feed, representative figures being included in the paper in tabular form.

No abstract