On Fri, 06 Jan 2006 11:18:25 +0000, Roger wrote:

The message
from John Stumbles contains these words:

....
... A single long rad with flow at one end and return at the other
will surely exhibit a similar temperature difference along its length (I
must test this when I get the 21374 I've just ordered from Toolstation :-))

However I can't see that joining multiple rads at top and bottom would
significantly change the heat distribution compared to connecting them
only at the bottoms (or, for that matter, at the tops).

The way I see it is that the method of connection does matter.
....
I think daisy chaining via both top and bottom connectors would
approximate to a normal radiator with the proviso that the restrictions
between the radiators might give a bigger temperature drop at the
junctions than would be expected had the top and bottom channels
continued with the same cross sectional area.

I agree with you now; and my 21374 does also: there seems little
temperature difference along the top of a rad so I think joining the rads
at the top as well as bottoms should tend to keep it that way.