LONDON Royal Society, March 5. G. C. ULLYETT: Host selection by Microplectron fuscipennis, Zett. (Chal-cididae, Hymenoptera). Microplectron was able to discriminate with ease between true and false hosts even where the latter resemble the normal host in everything except the presence of the living larva within the cocoon. The evidence seems to suggest that the acceptance of a host depends, to a large extent, upon the presence of larval movement. A new technique was developed, during these studies, which removed a number of objections present in previous methods. The selection within the host species as represented by choice between parasitised and healthy individuals and also between the latter and unsuitable hosts was investigated. A high degree of discrimination existed between healthy material and hosts containing parasite larvae which were well grown. The presence of parasite eggs, on the other hand, did not deter females from ovipositing in the host. A wholly mechanistic view of host selection is untenable; the underlying basis of behaviour is of a psychological nature. I. W. ROWLANDS and A. S. PARKES: A study of anti-thyrotropic activity. Inhibition of the effect of thyrotropic extract on the weight of the thyroid of the immature guinea-pig has been used as a test for anti-thyrotropic activity. The normal blood serum of the goat, horse, sheep, cow, and rabbit was not found to possess appreciable anti-thyrotropic activity, nor was that of a castrated ram injected with thyrotropic extract for four weeks. Anti-thyrotropic activity was induced in the blood of rabbits injected daily with thyrotropic extract over a long period. The activity began to appear after four weeks' injection and rose to a maximum at 10 weeks. 2 c.c. of serum obtained at this time completely inhibited the activity of an amount of thyrotropic extract otherwise sufficient to double the weight of the thyroid of the immature guinea-pig. A technique is described for the assay of anti-thyrotropic activity. G. A. MILLIKAN: The kinetics of muscle haemoglobin. The rates of reaction of muscle haemoglobin with oxygen and carbon monoxide have been measured by means of a micro-photoelectric form of the hartridge Roughton streaming fluid apparatus. The approximate velocity constants for extracts of norse heart muscle were compared with those for the blood haemoglobin from the same animal. The kinetic results indicate that muscle haemoglobin should be available as a naturally occurring intracellular indicator of oxygen tension, with a time lag of less than 1/100 second. This provides a new tool for studying the time relations of oxygen consumption in muscle. Its oxygen affinity, its concentration in muscle, and its rates of reaction are all such as to fit muscle haemoglobin for the role of an oxygen store which can tide the muscle over from one contraction to the next. No known property, however, precludes the possibility of its acting catalytically within the cell.