The discovery of cosmic radio emission by Karl Jansky in the course of searching for the source of interference to telephone communications and the instrumental advances which followed, have led to a series of new paradigm changing astronomical disco
veries. These discoveries, which to a large extent define much of modern astrophysical research were the result of the right people being in the right place at the right time using powerful new instruments, which in many cases they had designed and built. They were not the result of trying to test any particular theoretical model or trying to answer previously posed questions, but they opened up whole new areas of exploration and discovery. Rather many important discoveries came from military or communications research; others while looking for something else; and yet others from just looking. Traditionally, the designers of big telescopes invariably did not predict what the telescopes would ultimately be known for. The place in history of the next generation of telescopes will not likely be found in the science case created to justify their construction, but in the unexpected new phenomena, new theories, and new ideas which will emerge from these discoveries. It is important that those who are in a position to filter research proposals and plans not dismiss as butterfly collecting, investigations which explore new areas without having predefined the result they are looking for. Progress must also allow for new discoveries, as well as for the explanation of old discoveries. New telescopes need to be designed with the flexibility to make new discoveries which will invariably raise new questions and new problems.
Recent ammonia (1,1) inversion line data on the Galactic star forming region Sgr B2 show that the column density is consistent with a radial Gaussian density profile with a standard deviation of 2.75 pc. Deriving a formula for the virial mass of sphe
rical Gaussian clouds, we obtain a virial mass of 1.9 million solar masses for Sgr B2. For this matter distribution, a reasonable magnetic field and an impinging flux of cosmic rays of solar neighbourhood intensity, we predict the expected synchrotron emission from the Sgr B2 giant molecular cloud due to secondary electrons and positrons resulting from cosmic ray interactions, including effects of losses due to pion production collisions during diffusive propagation into the cloud complex. We assemble radio continuum data at frequencies between 330 MHz and 230 GHz. From the spectral energy distribution the emission appears to be thermal at all frequencies. Before using these data to constrain the predicted synchrotron flux, we first model the spectrum as free-free emission from the known ultra compact HII regions plus emission from an envelope or wind with a radial density gradient. This severely constrains the possible synchrotron emission by secondary electrons to quite low flux levels. The absence of a significant contribution by secondary electrons is almost certainly due to multi-GeV energy cosmic rays being unable to penetrate far into giant molecular clouds. This would also explain why 100 MeV--GeV gamma-rays (from neutral pion decay or bremsstrahlung by secondary electrons) were not observed from Sgr B2 by EGRET, while TeV energy gamma-rays were observed, being produced by higher energy cosmic rays which more readily penetrate giant molecular clouds.
