Under the assumption that white-light flares are caused by energetic particles penetrating into the photosphere (Svestka, 1970a;Najita and Orrall, 1970) the known number of protons needed for the white-light emission is used to obtain an estimate of the production of neutrons occurring at the same time. In the case of the white-light flare of 23 May, 1967, the peak flux of neutrons at the Earth distance had to exceed 3 neutrons/cm ~ s, thus being detectable in space. This maximum neutron flux reached the Earth as early as the time of the maximum phase of the flare in the Ha light. However, reasonable estimates show that flares associated with a detectable neutron flux should be fairly rare phenomena, maybe as rare as the white-light flares.
In the past years many experiments were carried out in order to detect neutrons coming from the Sun and from solar flares in particular. So far, all results of these experiments have been negative. Since, however, the measurements have rarely been made during periods closely following strong solar flares, one cannot exclude the possibility that the neutron flux could exceed the sensitivity limit of the receptors after some particular flares in which the production of neutrons was strong enough.
Lingenfelter et al. (1965) and Lingenfelter and Ramaty (1967) discussed theoretically the production of neutrons by energetic protons impinging into the photosphere, and estimated the peak flux of neutrons at the Earth after the major cosmic-ray flare of 12 November, 1960, to 33-70 neutrons/cm 2 s. This estimate was based on the meassured total flux of protons at the Earth distance from the Sun extrapolated to the solar surface, and on the assumption that the proton flux into space and towards the photosphere are equal. This procedure certainly involves great uncertainty in the deduced neutron flux.
Later on, Lingenfelter (1969) has estimated the neutron flux from solar flares under the assumption that the energy of optical emission in flares is completely provided by ionization losses of accelerated particles in the flare region. With this energy estimated to be 1029 erg/s and assuming an exponential rigidity spectrum of the accelerated particles in the flare region with Po = 100 MV, he has found that the neutron flux at the Earth should be slightly less than 10 neutrons/cm 2 s. This assumed energy transformation, of course, is of a hypothetical nature, and need not be true. The Balmer decrement in flares does not support this hypothesis, since it differs from a recombination spectrum (Kazachevskaya and Severny, 1958 ; gvestka, 1965), and in many flares there is no evidence at all for an acceleration process giving rise to suprathermal particles (cf., e.g., Svestka and Simon, 1969, and Svestka, 1970b). As De Feiter (1971) has pointed out, it is also difficult to understand why in this case the flare has a filamentary strue-* On leave from the Astronomical Institute of the Czechoslovak Academy of Sciences, Ond~ejov.