From analysis of the DIRBE weekly averaged sky maps, we have detected subst
antial flux in the 60 and 100 mum channels in excess of expected zodiacal a
nd Galactic emission. Two methods are used to separate zodiacal light from
more distant emission. Method I makes use of the time dependence of the nor
th-south annual variation observed at the ecliptic poles. This method is ro
bust against errors in the interplanetary dust (IPD) model but does not dem
onstrate isotropy of the background. Method II measures the ecliptic latitu
de dependence of the dust over a range of ecliptic latitudes (/beta/ > 35 d
egrees) at solar elongation e = 90 degrees. This allows the excess to be de
termined in each week of the DIRBE mission for high redundancy, but the res
ults depend weakly on the IPD model. Both methods give consistent results a
t 60 and 100 mum. The observed signal is consistent with an isotropic backg
round at the level nuI(nu) = 28.1 +/- 1.8 +/- 7(syst) nW m(-2) sr(-1) at 60
mum and 24.6 +/- 2.5 +/- 8 nW m(-2) sr(-1) at 100 mum.
The IR excess detected at 140 and 240 mum by these methods agrees with prev
ious measurements, which are thought to be the cosmic infrared background (
CIB). The detections at 60 and 100 mum are new. The integrated IR excess in
the window 45-125 mum is 23 +/- 8 nW m(-2) sr(-1), to be added to the 18 /- 4 nW m(-2) sr(-1) previously measured with the DIRBE and FIRAS instrumen
ts in the window 125-2 mm. While this new excess is not necessarily the GIB
, we have ruled out all known sources of emission in the solar system and G
alaxy. We therefore tentatively interpret this signal as the CIB and consid
er the implications of such energy production from the viewpoint of star fo
rmation efficiency and black hole accretion efficiency. However, the IR exc
ess exceeds limits on the CIB derived from the inferred opacity of the inte
rgalactic medium to observed TeV photons, thus casting doubt on this interp
retation. There is currently no satisfactory explanation for the 60-100 Clm
excess.