Opacity at Sendai was investigated following the lines of researches of Wright and Simpson. The obtained results are as follows:1. The values of opacity at Sendai toward N direction at noon are practically independent of relative humidity when the latter values are smaller than 50% but they increase rapidly with humidity when the values of humidity exceed 50%. This relation, is well explained if we assume, following Simpson, that change of opacity is due to acid nuclei which change their sizes with relative humidity. If, for instance, H2SO4 nuclei only are assumed as acid nuclei, we have, as their number and hygroscopic constant, N=190, Q=16×10-15 respectively. Or as the concentration of H2SO4 we have 3.04×10-11g/cc. However, the opacity-humidity relation at Sendai may also be explained by assuming that sea-salt nuclei as well as acid nuclei are effective to opacity. In this case, we have for NaCl, which is taken to represent sea-salt nuclei, N=85, Q=10×10-15 and the concentration of NaCl=5.1×10-12g/cc. and for H2SO4, which is taken to represent acid nuclei, N=130, Q=18×10-15 and the concentration of H2SO4=2.34×10-11g/cc.2. Next the relation between opacity and lapse rate of temperature is studied. The obtained result is the well established relation that opacity descreases according as lapse rate increases. This relation was hitherto explained that nuclei concentrate in the lower atmosphere in stable case causing bad visibility, and that they scatter high up in unstable air which cause good visibility, We have, however, shown that lapse rate and relative humidity have good negative correlation and in stable air nuclei have generally large sizes due to increased humidity. So that the increase of opacity in stable air will be caused not only by the increase of numbers of nuclei in the lower layer but also by the increase of sizes of nuclei. And it is shown that the latter effect is rather more predominant than the former effect in causing change of opacity with lapse rate.3. The relation between opacity and wind velocity was then examined and the well known relation, that opacity decreases with increase of wind velocity, was ascertained. It is, however, shown that for each group of data in which H>70% or H<70%, the values of opacity are independent of values of wind velocity. So that it may be concluded that number of nuclei effective for opacity in the air will not change appreciably with change of wind velocity in usual case. And the opacity-wind velocity relation obtained may ba explained by the change of sizes of nuclei with wind velocity, since wind velocity has good negative correlation with relative humidity.4. Opacity at each wind direction was then examined. It was shown that at Sendai opacity is large when wind blows from SE, or NE, or SW and it is small when wind blows from NW. Large opacity for SW wind will probably be due to the local effect of city of Sendai. And large opacities for NE and SE winds and small opacity for NW wind will be due to the effects of different air masses, which cover Sendai when respective wind is blowing. Because it is probable that Pm air mass corresponding to NE wind and Tm air mass corresponding to SE wind will contain large amounts of nuclei, whereas Pc air mass corresponding to NW wind will contain small amounts of nuclei.5. As to the seasonal variation of opacity at Sendai it was shown that opacity is generally large in summer having its maximum in July and is generally small in winter, its minimum occurring in October and in March intervened with secondary maximum of December.