We have found no paper on the detection of bursts near 21 cm of TZ Ari, Gl. 569, V371 Ori and VW Com. The four other stars had been detected at 21 cm before the observations of 1989-1993: YY Gem and EQ Peg reached flux densities of only a few mJy, which is detectable by the VLA, not by the Arecibo telescope, and, to our knowledge, only Wolf 424 and YZ CMi were detected in Arecibo at 21 cm (see Sect. 3.3 in Abada-Simon et al. 1994), but they are not reported in the tables of Appendix 1 because precise information is lacking. In addition, these latter four stars emitted no burst during numerous hours of observations in Arecibo near 21 cm, and these "negative" results have not been published (private communication). Table 3 (click here) gives a summary of the bursts noted down from the publications and a comparison with the 1989-1993 Arecibo campaign.
| Literature : | Literature : | 1989-1993 | 1989-1993 | |
| campaign: | campaign: | |||
| Stars | Number of | Number of | Number of | |
| observing | burst | observing | Number of | |
| hours at 1.4 | hours at 1.4 | bursts | ||
| GHz | GHz | |||
| Wolf 424 | 14 | 0 | 10.7 | 0 |
| TZ Ari | 0.4 | 0 | 3.3 | 0 |
| AD Leo | 31 | 6 | 38.3 | 12 |
| YZ CMi | 33.7 | 
| 7.7 | 0 |
| EQ Peg | 16.1 | >5 | 3.6 | 0 |
| GI 569 | 0.4 | 0 | 4.4 | 0 |
| YY Gem | 15.2 | 0 | 3.1 | 0 |
| VW Com | 0.4 | 0 | 2.0 | 0 |
Using the results of the literature to determine the parameter a for each star, and assuming that the probability to detect n bursts from one of the eight stars is estimated from Poisson's law, we deduce from Table 3 (click here) that during the 1989-1993 campaign in Arecibo we had a maximum probability (36.6%) of detecting one burst from EQ Peg, but a probability almost as high (32.7%) to detect no burst. On the other hand, we had the same probability (22.5%) to detect two or three bursts from YZ CMi, and only 5.1% chances to detect no burst. We therefore "should have" detected one or several bursts from YZ CMi; but observations of this star with the VLA show that it emits bursts of weak amplitude (a few mJy) over long timescales (a few hours), which is difficult to detect with the Arecibo telescope. Finally, AD Leo had a very low probability (5.6%) to emit the eleven bursts that we have actually detected. However, if the bursts detected in 1989-1993 actually correspond to less than eleven bursts, the probabilities inferred from the 1989-1993 campaign are then closer to those inferred from the literature. Concerning TZ Ari, Gl. 569, VW Com and V371 Ori, the published papers cannot allow to predict reliably the burst rate, since they report on less than half an hour of observation, which is not enough to establish significant statistics. Furthermore, Wolf 424 and YY Gem emitted no burst in 14 and 15.2 hours of reported observation, respectively: it is therefore not surprising that we detected nothing from them, but, on the other hand, the detection threshold published for Wolf 424 is much higher than that of the 1989-1993 campaign.
These Arecibo results can also be compared with some radio surveys of open clusters.
Bastian
et al. (1988) have used the VLA to search for radio emission from flare stars in the Pleiades:
they detected no emission at 1.4 GHz and their detection criteria led them to a frequency of "detectable"
radio emitting events of 
event in 120 hours; they concluded that the flare stars in the Pleiades may be
up to about ten times more active than flare stars in the solar neighborhood, i.e. they are not very
different. This seems to be confirmed if we put AD Leo at the Pleiades distance (125 pc): event its
strongest burst detected in Arecibo in 1993 would be below the detection level of the VLA survey by
Bastian et al. (1988). A survey of the Hyades made with the VLA at 1.5 GHz by
White et al.
(1993) led to similar conclusions within several uncertainties: the Hyades M dwarfs are not
much more active than the nearby population, and one can expect ý 1 event in 20 hours from the Hyades,
located at 45 pc. At that distance, one or two of AD Leo's strongest bursts could have been detected in the
40 observing hours made in Arecibo between 1990 and 1993, which gives a rate of detectable events
similar to the survey of White et al. (1993) or lower (for AD Leo's weaker bursts and for
the other stars, not detected during the Arecibo campaign).
Let us now compare the activity rates in radio with those in the visible domain. There is a report of about 0.8-1.5 optical flares per hour from AD Leo (Pettersen et al. 1984; Pettersen et al. 1986; Pettersen et al. 1990), to compare with about 0.3 radio bursts detected per hour in 1989-1993, and with about 0.2 radio bursts per hour predicted from the literature. Doyle & Mathioudakis (1990) estimate about 0.7 optical flares in 3 hr from YY Gem, whereas no radio burst was detected in the same amount of time in 1989-1993; with the same optical rate of YY Gem, one expects about 3 optical flares in 15 hrs, during which no radio burst was found in the literature.