The GEOS 1977 to 1981 observations published by Figer (1981b) show that the star had remained at a visual magnitude mv around 6.4 for three years (JD 2443500 to 2444700). Then a 0.4 magnitude light increase (from mv = 6.4 to 6.0) was detected in november 1980, followed by a rather rapid return to the previous light level, in the first months of 1981. The total event lasted 150- 200 days. The 45% light increase has been correlated with simultaneous spectroscopy showing the Balmer discontinuity in emission (Figer 1981b), which indicates a very violent atmospheric event. The corresponding light curve slopes are about - 4 mmag/day for the light increase (lasting about 60 days), and + 2.5 mmag/day for the decrease (lasting about 90 days), i.e. much slower than those observed in 1996 and described in the previous section.
A photographic monitoring of OT Gem has been carried out during 20 years at Sonneberg Observatory and published by Berthold (1983): it shows "slow irregular waves, superimposed by shorter variations". Given that this very extensive work was published in a rather difficult to find journal, we will briefly describe here the observed light behaviour.
During the years 1960 - 1980, the star had a mean photographic magnitude 
of
, for uninterrupted periods lasting between 300 and 800 days. This represents about 25% of the total observed time, and we will call it unexcited state.
Very often, however, the star showed light increases to 
(seven times in 20 years, or 12% of the time).
In two occasions these light increases were followed by a "plateau", that lasted for about 300 days.
Two more increases reached 
. They were followed by a slow -i.e. lasting around 1000 days- light decrease to the mean 6.10 value of . Once even the magnitude 5.7 was reached, this last value coinciding with the GEOS observations on
winter 1980-81.
During these same 20 years, there were two light decreases below the
6.10 mag level, to magnitude 
They
lasted about 130 days between the beginning and the end of the decrease.
So, although the light variations seem to occur randomly, a general behaviour above or under the mean magnitude of 6.10 can be summarized as follows:
Frequent 20% light increases that last from 40 to 170 days in total. The light curve can also reach a plateau that lasts about 300 days.
When a more violent -i.e. a 25%- light increase occurs, it is generally followed by a slow 
1000 days decrease.
The rare light minima - i.e. a light decreases of 20 to 25% - do not last more than 130 days. They only represent about 3% in the total time span of the observations, and do not seem to follow the light maxima.
We tried to define the time scales involved in these variations, and found the following:
The light increase rate is 
mmag/day
on the average. For the light increase observed at the
end of 1980 both at Sonneberg and by the GEOS,
we again get very similar values, respectively
-3.2 and -4 mmag/day, even with a very small number of available data.
The light decrease rate is generally 
mmag/day over a
few weeks. It can also be the slow rate already mentioned: 
mmag/day over 1000 days. These rates are much slower than those we observed in
1995-96 (Table 3 (click here)).
A remark should be made on the figures of Table 3 (click here). On the data of the previous observations, the time resolution is not as good as for those of 1995-96, so the rates of variation are very probably somewhat underestimated, but the qualitative difference between excited and unexcited state remains.
| Departure from: | Light increases | Light decreases | ||
| (in mmag/day) | (in mmag/day) | |||
| Unexcited state | ||||
| (during the years |  |  | ||
| 1960-80 and 1980-81) | (or  , "slow") | |||
| Excited state | ||||
| (plateau, |  |  | ||
| years 1995-96) | (peak at  ) |
(peak at  ) | ||