The diverse criticisms of jovian radio triggering, by Deach and Kaiser, are considered invalid primarily because they stemmed from ignoring the original event criteria. Moreover, attributing the observed triggerings to accidental encounters with a constant rotating beam is unjustified, as should have been obvious from the original analysis, since they occurred at the wrong jovian phases. On the other hand, there is nothing wrong with a rotating source also being triggerable, and it is proposed that the triggered emissions originate from just such sources.

Implying generation by natural radio lasers, the discovery of externally-triggered jovian radio emissions [Calvert, 1985a> was considered a breakthrough for understanding such planetary radiations. The observations of triggering, however, have been questioned by Desch and Kaiser [1985>, who found less statistical evidence for such events in the same Voyager radio observations, and this paper will answer their objections.

Although I respect Desch and Kaiser, in this instance I find them less than objective. They began with firmly-held doubts about triggering [Kaiser, private communication, 1981, to present>, and ever since have been casting about for some reason to reject the concept.

Their doubts seem rooted in a mistaken belief that triggering is bizarre and uncommon. The effect, however, is readily demonstatable with suitably-modified audio amplifiers, as was done before an audience at the past spring meeting of the American Geophysical Union [Calvert, 1985c>.

The radio triggering at Jupiter was considered a second instance of the same phenomenon previously found at the earth [Calvert, 1981, 1985b> and independently confirmed by Farrell [1985>. By attacking only the statistical significance of the triggering at only one planet, Desch and Kaiser have dealt with only part of the issue.

The principal evidence for triggering in the Voyager radio spectrograms was the sudden onset of certain specific features during incoming type III solar radio bursts, often simultaneously over a range of frequencies. An 'onset' in the original study pertained only to these specific spectral features, and although not explicitly stated, a prior absence of jovian activity was not considered essential (as in my Figure 3), since that would have favored occasions without suitable radio lasers [Calvert, 1982> pointing toward Voyager and waiting to be triggered.

The statistical significance of triggering was asserted only for the narrowband rising features ''having bandwidths less than 200 kHz, lasting twenty minutes or more, and rising at a few kilohertz per minute.'' Although broadband triggering was also pointed out, such events were completely excluded from my statistical analysis.

Desch and Kaiser, however, have ignored my event criteria and adopted new criteria of their ow, requiring the much less specific ''no jovian activity followed by sharply defined emission.'' They then ''construed'' triggering rather than simply trying to observe it. When this difference of criteria was pointed out to them, they subsequently ''tightened'' their criteria to include events ''like'' their Figure 1a (or my Figure 2). When asked specifically whether they had applied my criteria, their answer was ''no''. Their different results are attributed to this intentional choice of different event criteria.

In other words, although they purport to have examined my narrowband and broadband triggered events, they really have examined neither. Instead, they have simply examined something else and asserted that that was not statistically significant. Clearly, they have not tested the hypothesis they were seeking to disprove.

The difference of criteria can be illustrated with their own Figure 2, which shows brief triggering by a type III burst in the lower panel near 2I hr (although much compressed compared to the previous figures). Although hardly obvious at first glance, and difficult to verify without high-resolution spectrograms, that event came close to satisfyinb my criteria for narrowband onsets, since it occupied about 200 kHz, lasted roughly 15 minutes (four dots width), and seemed to drift upward at about 4 kHz/min. For Desch and Kaiser to have included such a case inadvertently clearly signifies that they have not honed their skills for finding such events. (This case also emphasixes the need to ignore prior activity, since the triggered feature was part of ongoing emissions detected in the other Voyager receiver channel.)

Their demonstration with artificial type III bursts signifies nothing extra, since it equally depended upon their original non-specific event criteria and yielded a spread which was consistent with Poisson statistics, given their stated mean. In the original analysis of the same data, the expected mean for accidental narrowband alignments was one event, as was the standard deviation (since the standard deviation of a Poisson distribution is simply the square root of the mean). This implies that the observed five coincident onsets were four standard deviations greter than the mean and hence quite significant. With their stated menas of 3.6 to 4.6, it would have required at least eleven events to have had comparble significance. It was wrong of them to imply that a few cases never matter when the standard deviations were different.