Based on observations with the James Clerk Maxwell Telescope JCMT, the Swedish ESO Submillimetre Telescope SEST and the Infrared Space Observatory ISO, an ESA project funded by Member States (especially France, Germany, The Netherlands and the UK) and with the participation of ISAS and NASA. The development and operation of ISOPHOT and the Postoperation Phase are supported by MPIA and funds from Deutsches Zentrum für Luft- und Raumfahrt.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

The extinction factors are exp{ $\tau$ $_{\rm\lambda}$ } and $\tau$ $_{\rm\lambda}$ /(1-exp{- $\tau$ $_{\rm\lambda}$ }) for the screen and the mixed case, respectively, with $\tau$ $_{\rm\lambda}=0.916\cdot A_{\lambda}$ (Genzel et al. 1998). For the same observed extinction curves, the amount of extinguishing dust has to be much higher in the mixed case, since the emitters at the shallow surface are practically not extinguished and those deeply embedded have to be more heavily obscured. Thus, e.g. A_V(KS- ${\rm screen}) \approx 60$ corresponds to A_V(KS- ${\rm mixed}) \approx 250$ . The length and orientation of the dereddening vectors in Fig. 1 remain practically unchanged, just the amount of extinction alters.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... properties

We used: $A_V:A_{\rm 7.7~ \mu m}:A_{\rm 15~ \mu m}:A_{\rm 25~ \mu m}:A_{\rm 60~ \mu m}:A_{\rm 100~ \mu m}:A_{\rm 850~ \mu m} = 1$ :0.039:0.039:0.036:0.015:0.0076:0.00035.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... $L_{\rm FIR}$

Because the MRN extinction curve shows features between 5 and 25 $\mu$ m, we used luminosities. $L_{\rm 10-40~ \mu m}$ comprises some kind of stable average over these features; our results are basically the same as when using e.g. the 12.8 or 15 $\mu$ m fluxes. The luminosities of the sample are derived from ISOPHOT and sub-mm photometry (Klaas et al. 2001); we use H₀ = 75 kms^-1/Mpc.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.