Brightness fluctuations at 1.65/x have been recorded by means of a 64-element array. Infrared photographs clearly show sunspots and granulation with a resolution better than 1". Quantitative analysis of the digitized data is used to compute autocorrelation and power spectrum. Half-width of autocorrelation (405 km) indicates a resolution comparable with the best observations in the visible range. Photographs and quantitative analysis show the existence of a strong contrast variation from the center to the limb. Seeing and instrumental effects are discussed. A model M.T.F. is utilized to compute a foreshortening correction. No attempt is made to get the actual absolute rms value. Nevertheless we find a definite variation of the observed rms which goes from 1.48 ~ :k ~z 0.15, at the center, to 1.05 ~ 4-0.15 at ~ = 0.7 (after foreshortening correction).
The solar atmosphere opacity reaches its absolute minimum at 1.65/t. This is the only region of the whole electromagnetic spectrum where observation of deep layers is possible. In optical depth a factor 2 is gained with respect to the visible region; this is of particular value for the study of the granulation, which is a display of a much deeper convection phenomenon.* Near infrared is also worth studying from another point of view: several authors (Coulman, 1972;L6na et aL, 1968) have predicted a significant improvement of atmospheric seeing with longer wavelengths. The question is still open and needs a serious study. Obviously a large aperture telescope is necessary in order to keep diffraction negligible (McMath Solar Telescope: 2/D ~ 0.2" at 1.65 #).
The observations have consequently been conducted with these two objectives: (a) accurate photometric data; (b) high resolution pictures. The last one has not been achieved as well as it might be, because the image quality of our infrared photographs is still limited by the scanning technique. So, the second question -does the seeing improve in infrared? -cannot be answered. Nevertheless we obtained numerous infrared 'photographs' which are very useful to check the results of the quantitative analysis. The observed fluctuations do really correspond to granulation and the strong center to limb variation of the contrast can be seen on every picture. This gives much more statistical weight to the results obtained from digitized records. * This factor 2 in the optical depth corresponds only to a height difference of 25 km. The question: "does the actual temperature fluctuations really change in such a small distance?" cannot easily be answered. Opacity effects very sensitive to the pressure distribution cannot be predicted at once. Furthermore, according to mixing length theory, the transition from radiative to convective transfer may take place in a distance as short as 20 kin. The present paper only reports observations; an interpretation in terms of temperature fluctuations will be given in another paper.