[morzh]

Ron

As I said, it IS in fact bypassed by the #29 cap to the "C" node. The readability of the schematic of course is not aided by putting a node in several places and then showing a decoupling cap in one of them, but then it is there.
It gets even better: while #29 is shown in the sch as a bypass for the output, it is physicall located at the coil #13 where the mystery cap is shown. But then it bypasses to "C" and not to the chassis which would make sense for the RF stages.
The output stage in turn has no decoupling other than the electrolytic capacitor (the #29 is in parallel to it).

The point of the correction is not just removing the cap from the sch (it is still there BTW when I went to the link above) but connecting the lower pin of the coil #13 (the one that is connected to the caps #11 and #12) to the "D" node. And removing that cap to the right of the resistor #47. Otherwise as you see we do not have the plate voltage connected to that coil (which is the load). And then the same "D" node is decoupled to the "C" by the #29 cap which is 0.05uF.

As for the mystery cap itself, the first time the question was raised in the thread the link of which I put above and obviously the cap was not there either. Could be a Philco wish list item. We even discussed a possible value then. It probably would not be a bad idea to put it there, of course having connected it to the "D" terminal.