By far the most likely answer to this question is that young elliptical galaxies are enshrouded in the dust formed by young stars, which absorbs the starlight, making them almost invisible at optical and near-infrared wavelengths. Figure 5 shows the effect of only a small amount of dust on the optical and UV light eminating from a young galaxy 1 Gyr old. The absorbed energy simply gets re-radiated as thermal emission giving rise to substantial emission at far-infrared and sub-millimetre wavelengths.
|Figure 5. The spectral energy distribution of a galaxy 1 Gyr old forming stars at a rate of 1 M yr-1. The top curve is the prediction with no dust (as in Figure 3) and the bottom curve the prediction assuming a simple dust creation model.|
In a parallel way to that of near-infrared astronomy some 10 years ago, recent technological developments in the last couple of years have led to the first searches for dusty PGs: SCUBA (the Submillimetre Common User Bolometer Array) was built by the Royal Observatory Edinburgh for use on the James Clerk Maxwell Telescope (JCMT) in Hawaii. It is the world's largest submillimetre camera, and takes images simultaneously at two submillimetre wavelengths. It can take deeper images of the sub-millimetre (450 µm- 2.0 mm) sky than ever before, mapping areas in minutes which previously used to take hours to complete and is quite capable of detecting re-processed starlight from dusty PGs at high redshift.
Fueling the excitment surrounding the anticipated arrival of SCUBA was the report of the first detection of an extragalactic far-infrared background from the COBE satellite (Puget et al. 1996). This background appears isotropic over the range 400 µm- 1.0 mm and has a natural explanation in terms of re-radiated starlight from dusty galaxies at high redshift (see Puget et al. 1996). Despite only begining regular observations in May 1997, SCUBA has not proved disappointing: detections of sources with submillimetre spectral properties consistent with high redshift dusty PGs were reported last year (Smail et al. 1997). Some caution is necessary in the early interpretion of these results since the Smail et al. search fields cover only a few arcmin2 of sky and were not the traditional blank fields characteristic of previous PG searches, but instead were centered on known distant clusters of galaxies. This was done in order to utilise the familiar gravitational lensing effect seen in rich clusters, which amplifies the light from backgound objects as it passes through the cluster's gravitational potential well. While this makes distant galaxies brighter and therefore easier to detect, source luminosities and surface density estimates require accurate knowledge of the amplification factor, which is intrinsically uncertain.
Very recently, the results from two independent blank field searches have been reported (Hughes et al. 1998, Barger et al. 1998). Each group surveyed approximately 10 arcmin2 of sky (including the celebrated Hubble Deep Field). These data reveal objects with submillimetre spectral properties consistent with that expected from dusty galaxies in the redshift range z = 2-4, powered by star formation with an implied SFR 5-10 times larger than the Lyman dropout galaxies seen in optical surveys. Results from surveys covering larger areas of sky are eagerly awaited: if these and similar experiments confirm a dusty PG population, the last empirical ingredient required to understand the formation of galaxies may have been found and a description of galaxy evolution over a wide span of cosmic time would then be within reach.