NavList:

On Jan 27, 2006, at 9:45 PM, Ken Gebhart wrote: 
> Jim, 
> 
> Here is my take sextant certificates. .... 
 
> 1. In addition to the regular 10 deg correction values, correction 
> values for each in between degree (which you are not shown) can go 
> quite far from expected values. For example on one sextant, for 
> 40, 50, and 60 degrees, we have corrections of +10, +9, and +2 
> secs respectively. Not bad. But at 47 and 54 degs we have a 
> correction of 0! So, you can see it is not a straight line 
> variation between the 10 deg values. 
 
Thanks for sharing this with us. The one place where I have seen 
data like these is for late-model Husun sextants available for sale 
on eBay, where a photo of a British Admiralty certificate is 
inclu! ded. The certificates include a continuous curve, although I 
don't know whether that represents continuous data or interpolation 
between set measurement points. On these certificates you can see 
the errors wandering considerably between the standard correction 
points. 
 
> So, if you combine all of these uncertainties, it makes the 
> accuracy certificate something to be taken with a grain of salt. I 
> have a SNO-T sextant with a certificate that says straight zeros 
> across the board. Of course this is ridiculous. Whenever I see a 
> claim by a manufacturer of less than +/- 15 sec. I believe they are 
> making a marketing statement rather than a true one. 
 
The larger Husuns were the "luxury" models compared to the "Mate." 
Oftentimes these would show zero or near zero corrections, both on 
National Physical Laboratory and on Husun! certificates, whereas the 
Husun certificates issued with "Mates" usually show errors of up to 
50 seconds. Because of this, I believe the certificates for the 
large Husuns may not have been lying too awfully much. 
 
An optical physicist described a few years ago on the Yahoo sextant 
list his testing of his Astra sextant. He said it was very good. 
 
Fred Hebard.