The effectiveness of large telescopes is limited by the terrestrial atmosphere rather than by their optical or mechanical performance. Due to its refractive, absorbing and scattering properties, the atmosphere is one of the important elements that determines the limiting magnitude, resolving power and the wavelength range within which observations are possible. These effects are present even on the clearest nights. In addition, general meteorological conditions such as cloudiness, wind, high humidity, or extremely low temperatures further impair or prevent the operation of astronomical equipment. These effects can be avoided by placing the equipment above most or all of the atmosphere (balloon-borne telescopes, satellite observatory, lunar observatory). For the foreseeable future the cost of placing and operating a telescope outside the atmosphere will be such that it will be restricted to the investigation of problems that cannot be solved with observations made only from the ground. It seems possible, however, to choose sites for terrestrial observatories where the atmospheric effects are minimized. Just where the optimum conditions can be found is not yet known, but it is highly probable that none of the existing observatories is located on the World's best site. In fact, it is quite possible that conditions may be found on hitherto unknown sites that could bring a considerable advance in astronomical research before going to space equipment. Whether or not it is practicable to erect and operate an astronomical observatory on the optimum site (or sites) will depend largely on location, because in addition to excellent astronomical observing conditions the site has to be accessible and capable of development for satisfactory working conditions.

Les ondes lumineuses émises par les astres doivent nécessairement traverser l'atmosphère terrestre avant de parvenir aux instruments d'observation astronomique. Au cours de cette traversée, divers phénomènes atmosphériques affectent ces ondes aussi bien en ce qui concerne l'énergie qu'elles transportent qu'en ce qui concerne leur forme ou leur orientation. Il surgit ainsi pour l'astronome une série de problèmes particuliers, consistant principalement dans la manière d'éviter au maximum les effets des phénomènes atmosphériques, effets pouvant être désastreux pour la qualité et le rendement des programmes d'observation.

Rösch. — We have already spent one third of the time available for this Symposium and I think it would be useful to say a few words about the present situation as I see it. We have had very lively discussions, but I am sure most of you have the feeling that there has been some confusion. This indicates that everyone is interested in the subject, but I hope that now we can restrict the discussion to better-defined fields. My feeling is that among all the factors involved that of turbulence is by far the most difficult. However we still have problems regarding the meteorological elements other than turbulence. I think it is excellent that the meteorologists here have been interested, although not always in agreement with the astronomers (but to disagree is much better than to have no ideas). What I would like to suggest now is that, before the end of this meeting, Dr. Dommanget have private discussions with the meteorologists about how we can use, for preparing site testing programmes, the data already collected by the meteorologists, and perhaps other data which they will give us in the course of their work. I hope that Dr. Dommanget can give us a 10-minute summary of these problems in one of the last sessions.

The concern has been expressed many times by Dr. Bowen and others that a significant portion of the seeing deterioration may occur in levels of the atmosphere very near the ground, within a few tenths of meters of the ground. When I refer to the quality of seeing I am refering to the image size one observes in a telescope of very large aperture and I will assume that this is equivalent to image motion as observed in telescopes of very small aperture. I will not attempt a further justification for this concern; however this is the basis for the studies we are just beginning at Kitt Peak, where we will attempt to quantitatively show whether or not there is need for concern about the very low levels of the atmosphere. So we begin with the thesis that much of the poor seeing observed at a site, the enlargement of photographic or visual images as observed through a large telescope, is due to refractive inhomogeneities in the lower levels of the atmosphere, within less than 100 m above the telescope. We presume that these inhomogeneities are of local origin and that their distribution and motion is determined primarily by site topography, wind direction and velocity. The few experiments we have made thus far at Kitt Peak have been designed to ascertain quantitatively the importance of these factors. Our approach has been to make observations of the large-aperture seeing with simultaneous observations of the thermal structure of the air accessible to us immediately above the telescope.

There are several different atmospheric causes, of imperfections in the images seen in a telescope. The most serious are due to thermal convection from the ground, or the stirring of a stable layer at the ground by wind : choice of site and observatory design can reduce these effects. In the free air well away from the ground some of the largest inhomogeneities may be due essentially to large gradients of humidity, and certain cloud types indicate the mechanisms by which such inhomogeneities are established. Temperature fluctuations caused by pressure fluctuations in eddies close to the telescope are probably not serious, and can be avoided by observatory design.

Dans cet exposé, on cherche à définir les conditions que doivent remplir les instruments utilisables en campagne pour une mesure aussi bonne que possible des paramètres caractéristiques permettant de choisir l'emplacement d'un grand télescope d'au moins 3 m de diamètre.

While at night almost all optical effects of the atmosphere have to be investigated by using stellar images, the solar observer has at his disposal the solar disk of about half a degree in diameter, covered all over with low contrast details of about 1″, the granules, and confined to the sky by a sharp limb. This enables him to observe all effects occuring within the area of the Sun simultaneously, to follow even the propagation of certain atmospheric effects across the disk. On the other hand he has only the Sun and cannot — as night astronomers might do — extend his observations to all directions in the sky.

Ramberg. — According to the agenda, the discussions this morning are to be concentrated on field surveys. We are very sorry that our Russian colleagues from the Pulkovo Observatory, Drs. Melnikov and Kutcherov, have not arrived; I am sure that all of us would have been extremely interested in hearing about their experiences. We are glad, however, that other astronomers with great experience of field surveys and seeing expeditions will be able to tell us about site testing in Southern Africa, Chile and Australia.

I have not prepared anything on the subject of the present session.

The best I can do is to take as an example the organization of the field survey in Southern Africa, by the European Southern Observatory (E. S. O.) with which I was collaborating from November 1955 to March 1957.