High statistics search for ultrahigh energy γ-ray emission from Cygnus X-3 and Hercules X-1

We have carried out a high statistics $(2\mathrm{}\times {10}^9$ events) search for ultrahigh energy γ-ray emission from the x-ray binary sources Cygnus X-3 and Hercules X-1. Using data taken with the CASA-MIA detector over a five year period (1990–1995), we find no evidence for steady emission from either source. The derived 90 % C.L. upper limit to the steady integral flux of $\gamma$ rays from Cygnus X-3 is $\Phi (E>\mathrm{}115\mathrm{}$TeV)$<6.3\mathrm{}\times {10}^{-15}$ photons ${\mathrm{cm}}^{\mathrm{-}2}$ sec${}^{-1}$, and from Hercules X-1 it is $\Phi (E>\mathrm{}115\mathrm{}\mathrm{TeV})<\mathrm{}8.5\mathrm{}\times {10}^{-15}$ photons cm${}^{-2}$ sec${}^{-1}$. These limits are more than two orders of magnitude lower than earlier claimed detections and are better than recent experiments operating in the same energy range. We have also searched for transient emission on time periods of one day and $0.5$ h and find no evidence for such emission from either source. The typical daily limit on the integral $\gamma$-ray flux from Cygnus X-3 or Hercules X-1 is ${\Phi }_{\mathrm{daily}}(E>\mathrm{}115\mathrm{}\mathrm{TeV})<\mathrm{}2.0\mathrm{}\times {10}^{-13}$ photons cm${}^{-2}$ sec${}^{-1}$. For Cygnus X-3, we see no evidence for emission correlated with the $4.8$ h x-ray periodicity or with the occurrence of large radio flares. Unless one postulates that these sources were very active earlier and are now dormant, the limits presented here put into question the earlier results, and highlight the difficulties that possible future experiments will have in detecting $\gamma$-ray signals at ultrahigh energies.