The relationship of oviposition to weather, with a standard stimulus (4 per cent. ammonium carbonate), has been tested by a series of 134 exposures of a group of three sheep for one hour. 190 egg clusters were obtained, of which 169 were identified. With the exception of one cluster of L. caesar all were L. sericata.

There was a fairly close correlation (0·586) of oviposition with air-temperature. When the possible effect of other factors was eliminated, the correlation was little affected, showing that temperature was in itself responsible for the variations in oviposition. The lower critical limit of temperature was in the neighbourhood of 56°–58°F.

Solar radiation and oviposition had an apparent correlation, but this was found to be spurious, the effect being due to related air-moisture changes. At temperatures near the lower critical limit, insolation did not increase the oviposition rate, or lower its temperature threshold.

Humidity had an apparent inverse correlation with oviposition, but this also was an indirect effect, due to related temperature changes. Humidity showed a closer correlation when expressed as saturation deficiency than as relative humidity. Evaporation rate from a white atmometer sphere had only a low correlation with oviposition.

It is shown that the partial correlations of both saturation deficit and insolation with temperature effects eliminated, are insignificant, i.e. their apparent effects are due to related temperature changes. The multiple correlation coefficient for temperature humidity and insolation is practically identical with that for temperature.

Independent variations of insolation, i.e. variations not simultaneous with related temperature changes, had a significant effect on oviposition; for a given temperature, a high intensity of insolation was less favourable than a low, and this appeared to be a true insolation effect, and not due to associated humidity changes.

Independent variations of humidity had a suggestive but statistically insignificant effect, humidity tending to be inversely related to oviposition.

The regression of log oviposition on temperature, within the limits of this experiment, i.e. at temperatures up to 72°F., was linear. The slope was not significantly affected by the insolation and humidity factors, the total coefficient being 0·, and the partial coefficient 0·0428. These coefficients are the increase of the log number of egg clusters (plus one) for 1 degree F. increase, i.e. the oviposition rate increased in geometric progression for unit increase of temperature, the value (N+l) for egg clusters doubling itself for 7° increase.