In this study we report the first observation of spontaneous Raman solitons in stimulated Raman scattering (SRS) by the gas NH 3. The scattered radiation is called Stokes radiation. Raman solitons are of considerable interest, because their existence can be explained by quantum-mechanical fluctuations of the electromagnetic field in vacuum. We have observed spontaneous Raman solitons in a forward SRS configuration for two different molecular transitions of NH 3, the laser emissions at 58 μm and 72.6 μm wavelength. These are optically pumped by 10 μm CO 2-laser pulses with a duration of 100 ns and an energy of 150 mJ. Spontaneous Raman solitons are short spikes in the pump pulse which occur during its depletion. Their origin is the rapid π phase change of the Stokes seed. In contrast to other laboratories we have used single-pass cells. Thus, we have succeeded in observing multiple spontaneous Raman solitons during one pump pulse. Previous experiments with multi-pass cells never showed multiple solitons. Since multiple spontaneous Raman solitons have already been reported in an earlier experiment with a single-pass cell filled with hydrogen at high pressure, we conclude that such multiple Raman solitons can be observed mainly in this type of gas cell. Subsequently, we have performed statistical measurements on the delay time and the height of the spontaneous Raman solitons in the depleted pump pulse for the 58 μm-NH 3 emission. We have compared these statistics with theory and equivalent experimental results of other laboratories. They are in good agreement with the assumption that quantum-mechanical fluctuations are the origin of spontaneous Raman solitons. The most recent theories postulate that the origin of the formation of spontaneous Raman solitons can be explained by the rapid π phase change of the Stokes seed as well as that of the laser or polarization wave. Therefore, we have determined the phase of the spontaneous Raman solitons relative to the depleted pump pulse. Although, such changes of sign of the relative phase have already been observed in an earlier SRS experiment with hydrogen at high pressure, we did not detect any in our experiment. Therefore, we conclude that in this experiment the π phase change occurs in the Stokes or polarization wave.