4 Conclusions

1. We have investigated the evolution of the H₂O maser emission in G43.8-0.1 during a series of flares that happened at radial velocities of 37.5-39.5 km s^-1 in 1996-1998. For the strongest line of this time interval, the one at $V_{\rm LSR}=38.2$ km s^-1, we have found that the flux density and linewidth are related as $\ln F\propto\Delta V^{-2}$. This indicates that the maser in this condensation is unsaturated or partly saturated. The flare at 38.2 km s^-1 took place near the minimum of activity of the G43.8-0.1 maser and started a new activity cycle of the H₂O maser in this source.
2. We have traced the descending branch of evolution of the 42.2 km s^-1 feature. We have got a complete history of evolution of this long-living component during 16 years. The initial and final stages of the evolution turned out to be identical in duration, character of the flux and radial-velocity variability, being mirror images of each other. The only difference is that in both cases the line was narrowing (from 0.70 to 0.57 km s^-1 at the initial stage and from 0.55 to 0.48 km s^-1 at the final). Line narrowing, accompanied by a decrease of the flux at the descending branch of the evolution could be a consequence of a change of some parameter of the medium hosting the 42.2 km s^-1 maser emission. For instance, temperature could change. The activity of this maser condensation correlated with that of the maser at whole, but always with a time delay of about one year.
3. The correlation between flux variation and linewidth is a common phenomenon in G43.8-0.1. This phenomenon, taken together with the existence of a long-living feature at 42.2 km s^-1, testifies to the fact that the regions responsible for the H₂O maser emission (maser condensations) in G43.8-0.1 are sufficiently stable structures. In this case, the envelope may be quite clumpy, and minor changes of the pumping conditions can considerably increase the emission intensity, provided the maser is partly unsaturated. Such stable structures can exist, for instance, in the case of the presence of a molecular outflow originating in the region of formation of a massive star.
4. The time interval between the minima of maser activity in G43.8-0.1 was 18 $\pm$ 1 yr. This figure can be taken as the period of the maser activity and is possible of the central-star activity in G43.8-0.1 in the process of its formation.

Acknowledgements

The author is grateful to the staff of the Pushchino Radio Astronomy Observatory for great help with the observations.