This paper has been difficult to review because it is written in stilted english verging on pidgin. I strongly recommend that the authors ask a native english speaker to read it prior to / resubmission. The data however are quite interesting, and could be published as useful background to a more analytical study. The use of the perforated patch technique to look at these otherwise labile responses is very nice. It should be made clear that the series resistance measured on page 6 (3 Iines from bottom) reflects the access resistance through the perforated patch.

Scientific Points

Is the 32% of Icaffeine left in TEA sensitive to apamin?

Is it sensitive to norepinephrine?

　It is not clear whether in an individual cell the response to increasing amounts of caffeine increases smoothly or whether there is a threshold obscured by pooling data from different cells (Fig. 4).

Is I Icaffeine affected by charybdotoxin?

Pg. 11 Iine 4 is misleading, since the relation between [Ka]i and Icaffeine was only very indirectly investigated, using A23187 and ryanodine.

Pg. 12 Iine 4, note this solution also contained EGTA. Also, in Fig. 6A, illustrate a partly recovered response (is its shape conserved?).

Pg. 13, bottom. From whom was the information about frontal cortex received?

Pg. 16. Fig. 16 seems to show that rather than no recovery in zero Ca, there is a fast component (~20%) of recovery which doesn't depend on [Ca]i.

It would have been nice to test the effect of ruthenium red.