Recently correlation analyses between Galactic dust emission templates and a number of CMB data sets have led to differing claims on the origin of the Galactic contamination at low frequencies .
de Oliviera-Costa et al ( 1999 ) have presented work based on Tenerife data supporting the spinning dust hypothesis .
Since the frequency coverage of these data is ideal to discriminate spectrally between spinning dust and free-free emission , we used the latest version of the Tenerife data , which have lower systematic uncertainty , to study the correlation in more detail .
We found however that the evidence in favor of spinning dust originates from a small region at low Galactic latitude where the significance of the correlation itself is low and is compromised by systematic effects in the Galactic plane signal .
The rest of the region was found to be uncorrelated .
Regions that correlate with higher significance tend to have a steeper spectrum , as is expected for free-free emission .
Averaging over all correlated regions yields dust-correlation coefficients of 180 \pm 47 and 123 \pm 16 \mu K /MJy sr ^ { -1 } at 10 and 15 GHz respectively .
These numbers however have large systematic uncertainties that we have identified and care should be taken when comparing with results from other experiments .
We do find evidence for synchrotron emission with spectral index steepening from radio to microwave frequencies , but we can not make conclusive claims about the origin of the dust-correlated component based on the spectral index estimates .
Data with higher sensitivity are required to decide about the significance of the dust-correlation at high Galactic latitudes and other Galactic templates , in particular H _ { \alpha } maps , will be necessary for constraining its origin .