4 Summary

We find for the IRAM 30-m telescope that

(1) there is agreement within 10 - 15$\%$ between the rms-values $\sigma$$_{\rm g}$($\epsilon$) derived from the radiometric measurements and the finite element calculations (see Fig. 1). This accuracy is set by the radiometric measurements and the atmospheric correction, rather than by the accuracy of the FE structural calculations.

(2) Figure 3 indicates that the on-axis gain-elevation dependence G₀($\epsilon$,$\epsilon$₀) holds for extended sources not exceeding in diameter approximately two half-power beamwidths ($\theta$$_{\rm S}$ 2$\theta$$_{\rm b}$), i.e. approximately the diameter of the main beam. This result implies that the gain-elevation dependence can be determined without correction from the measurement of extended sources, for instance the planets, in case they do not exceed 2$\theta$$_{\rm b}$ $\equiv$ $\theta$^*. Table 1 gives the values $\theta$^* and largest planet suitable for determination of the gain-elevation dependence. As evident from Fig. 3, the gain-elevation dependence decreases for sources larger than two beamwidths and disappears for sources larger than $\sim$8 beamwidths.

(3) for any wavelength in the range 0.8 mm $\lambda$ 3 mm, the on-axis gain-elevation dependence G₀ is obtained from Eq. (3) using the radio-effective rms-values R$\sigma$$_{\rm g}$($\epsilon$) of Table 2. The gain-elevation dependence for extended sources is obtained from these values and the data of Fig. 3, using Eq. (9).

Acknowledgements

The homology calculations were originally made by ARGE KRUPP (now VERTEX) and MAN, Germany, and later repeated by P. Raffin (formerly at IRAM) and M. Bremer (IRAM). We profitted from many discussions with our colleagues in the effort to measure a reliable gain-elevation curve. We thank the referee, Mr. J. Baars, for his comments and his Occam razor.