A complete investigation of the variations likely to occur in the lengths of larvae of Heterodera rostochiensis (for example) under all conditions would involve measuring larvae from a great number of varieties of potato grown in a great number of different localities, and also of larvae from tomatoes of different varieties grown under various conditions. The results obtained so far for this species of Heterodera make but a small contribution towards the mass of information which might be accumulated. Until such further data become available, certain conclusions have been drawn as to the best method of taking samples of larvae for measurement, but it must be remembered that these conclusions may need revision at a later date in the light of new information. According to our present knowledge it appears that the variety of potato on which a cyst has developed does not significantly affect the mean length of the larvae within, but that the mean length of the larvae from a single plant, irrespective of the variety, may differ significantly from that of larvae from another plant. It is also deduced that the variations in the mean lengths of the larvae from different cysts are greater than the variations within the individual cysts. From these facts it follows that if one wishes to estimate parameters for larval lengths of a population of H. rostochiensis larvae, a better estimate will be obtained if one takes a small number of larvae from each of a large number of cysts rather than the same total number of larvae from a small number of cysts. Similarly the cysts should have come from a large rather than a small number of plants, but apparently it is not necessary that the plants should be of as many varieties as possible, though this last assumption is based on information derived from only seven varieties of potato, all of them maincrop varieties, and it may need modification when larvae from other varieties have been examined.

An analysis of slaughter-house returns, covering over 73,000 cattle slaughtered during three weeks in June, 1942, reveals evidence of heterogeneity and interaction in respect of all the five categories involved: time, area, class, origin and degree of infestation with liver fluke. By heterogeneity is here meant that the differences between the various sub-divisions of any one category are too large to be ascribed to chance; by interaction is meant that the relative proportions between the subdivisions under any one category do not remain steady (within the limits of random variation) over the co-ordinate subdivisions of a second category. This makes the data difficult to interpret clearly.

The action of Trichinella spiralis infections in markedly lowering protein digestion in rats is confined to a period 4 to 12 days afier infection. Minor disturbances in protein digestion occurring about 30 days after infection may have been due to adult females leaving the host.

Over the first 12 days of infection the protein digestion was lowered (calculated as the summation of the percentage fall for each of the first 3 periods) approximately 1% for each 100 infective larvae fed. The degree of digestive disorder seemed unrelated to the number of larvae finally recovered from the hosts' tissues.

Inorganic P excretion fell to a very low level 4 to 8 days after infection. Twenty-four days later the rate of excretion rose, the maximum being 2 or 3 times the normal level.

Ca assimilation fell to a very low point during the second period of infection, falling again after a further 9 periods.

As indicated by weight losses the most marked period of the disease occurred from the 8th to the 11th period after infection.