We report the serendipitous detection of a significant overdensity of Herschel -SPIRE 250 \mu m sources in the vicinity of MRC 1138-26 .
We use an adaptive kernel density estimate ( AKDE ) to quantify the significance , including a comparison with other fields .
The overdensity has a size of \sim 3 \aas@@fstack { \prime } 5 - 4 \arcmin and stands out at \sim 5 \sigma with respect to the background estimate .
No features with similar significance were found in four Herschel -observed extragalactic control fields : GOODS-North , Lockman , COSMOS and UDS .
The chance of having a similar overdensity in a field with the same number but randomly distributed sources is less than 2 % .
The clump is also visible as a low surface brightness feature in the Planck 857 GHz map .
We detect 76 sources at 250 µm ( with a signal-to-noise ratio greater than 3 ) , in a region of 4 \arcmin radius ( or 1.51 \pm 0.17 arcmin ^ { -2 } ) ; 43 of those ( or 0.86 \pm 0.13 arcmin ^ { -2 } ) are above a flux density limit of 20 mJy .
This is a factor of 3.6 in excess over the average 0.24 \pm 0.02 arcmin ^ { -2 } in the four control fields , considering only the sources above 20 mJy .
We also find an excess in the number counts of sources with 250 µm flux densities between 30 and 40 mJy , compared to deep extragalactic blank-field number counts .
Assuming a fixed dust temperature ( 30 K ) and emissivity ( \beta = 1.5 ) a crude , blackbody-derived redshift distribution , z _ { \mathrm { BB } } , of the detected sources is significantly different from the distributions in the control fields and exhibits a significant peak at z _ { \mathrm { BB } } \approx 1.5 , although the actual peak redshift is highly degenerate with the temperature .
We tentatively suggest , based on z _ { \mathrm { BB } } and the similar S250/S350 colours of the sources within the peak , that a significant fraction of the sources in the clump may be at a similar redshift .
Since the overdensity lies \sim 7 \arcmin south of the z = 2.16 Spiderweb protocluster MRC 1138-26 , an intriguing possibility ( that is presently unverifiable given the data in hand ) is that it lies within the same large-scale structure .