From: Philo, John <jphilo@amgen.com>
  To  : 'rasmb@bbri.harvard.edu' <rasmb@bbri.harvard.edu>
  Date: Sat, 14 Sep 1996 14:54:25 -0700

RE: six channel centerpieces

Julius,

I thought one of the Beckman folks or Model E "old-timers" would respond
to your queries, but since they apparently haven't, I will tell you what
I know.

> 1) Can we still use the old six channel centerpieces from the model E, 
>> just put them in the new cell housing with beckman catalog number 334784,
>> or do we also need to buy new centerpieces.

No, you cannot use your old centerpieces with the new housings.  The new
housings have an extra internal ring which holds the centerpiece in
place.  This feature allows you to tighten the bottom window against the
centerpiece with enough force to maintain a seal while you fill the
centerpiece from the top.  The old "E" centerpieces do not have a step
on their outside to mate with that ring (or at least that is true of the
ones I have).

The new centerpieces have the channels moved radially inward.  If you
use the old ones channel C extends to too high a radius and you cannot
get light through all the way to the bottom of that channel.

>> 2) Persons using those new cells with the six hole centerpieces are they
> satisfied about the obtained results with those cells, specially the
>>third > compartment?

Channel C works okay with the new cells, at least on my instrument.  But
you should be aware of two things that affect the quality of the data
near the extremes of the radial range on the XL-A.  First, even with the
new centerpieces and housings, the window holders still partially
obscure the very top of channel A and the very bottom of channel C. 
Some XL-A users have manually modified the window holders to take care
of this.  Second, depending on how the lamp is positioned, the incident
light intensity can fall off dramatically at the extremes of the radial
range.  In my instrument the intensity if typically 30-40% less at 7.2
cm than at 5.8.  This, combined with the window holder blockage, means
that I do usually see a fall-off in the quality of the data at the
bottom of channel C.  For this reason, I usually try to set up the
experiment so the lowest sample concentrations are in channel C.

'Best,

John Philo, Amgen