IntroductionThe significance of periodic fluctuations and/or a continuous spectrum of fluctuations of the solar thermal radio emission in the absence of solar flares is doubtful. BENZ and FiiRs~ (1987) used the large aerials of Arecibo and Effelsberg and did not find any discrete periodic oscillations of coinciding period in the records, nor they have found discrete signals in the cross correlation of both records. A positive answer to the question "Are radio fluctuations real?" would have serious physical implications (relevance to coronal heating).
The observing programWe report on a coordinated fluctuation observing program between the Riga Radioastrophysical Observatory (RO) and Tremsdorf Solar Radio Astronomy Observatory (OSRA). Both stations are abopt 850 km apart. The decimeter range was chosen with respect to the present technical facilities. We used in our observing program an altazimutal mouqted 10 m parabolic mirror with focal ratio 0.43 in RO, driven by a coordinate transformer on rotating transformers. At OSRA, an equatorially mounted 7.5 m parabolic mirror with a focal ratio of 0.25 was applied, driven by a quartz clock. The halfpower beam width of the aerials is approximately 3.5 deg. The two identical receivers were built at RO working at a central frequency of 755 MHz. The bandwidth is 5 MHz, output time constant 1% this means AT/T = 1.2 .To minimize the influence of receiver gain instabilities the difference between the signal from the aerial and an artificial noise source is recorded. At the beginning of observations the signal from the noise source is compensated with the signal from the aerial. Further the receiver gain is enhanced by 20-60 times the usually practiced level in routine observations of solar radio emission. The data have been digitally recorded using a sampling rate of 12 s. Between 1987 October 16 and November 27 there were 15 days with succesfully coordinated observations suited for studying fluctuations. These were days with stable weather conditions at both observing sites and low solar burst activity at least in the time interval to be compared. Generally not much more than two hours of the daily records are used for comparison.
Data analysisFirst, the time series of both observatories have been independently treated using the Maximum Entropy (ME) method of power spectrum estimation (cf. AURAS et al. 1984). On 12 days there is at least one coinciding spectral peak in the data. We have recognized three main problems : 1 . Several discrete spectral peaks on a background of coloured noise are present in the power spectrum of both records on most days.
