When the Naval Observatory overhauled a sextant they pasted a Certificate of Inspection to the inside of the box lid. When the N.O. serviced 3500/1542 they should have pasted such a certificate to the box it was in. The box found near the bones roughly two years later had no such certificate or the Brits would have known it was an American box instead of speculating about it being English or French. The inside of the lid is a protected location even in the harsh Niku environment. That’s a strong argument that the sextant that had been in the box found with the bones was not the sextant assigned to USS Bushnell.
<sigh>
This happens all the time.
I don't think I have ever kept track of data and not wished halfway through that I had been paying attention to more elements than I did at the beginning.
I did not create a column to keep track of which boxes have certificates and which do not.
Many of the discrepancies between the N.O. number engraved on the sextant and other numbers on the box come from the extant certificates, but it is not easy to extract the number of boxes-lacking-certificates from the list.
I don't have time this morning and may not have any time soon, but I am moderately confident that the vast majority of the entries I have made in the table come from people posting in the Forum thread,
"Can you add to the list of sextant numbers?".
Ah! Now that I look at the fields that I created for the table, one is for "Inspection date."
|-
! Maker
! Maker No.
! Navy No.
! Inspection date
! Comments
! Bubble
! N.O. # on box
So we can be pretty confident that the presence of an "Inspection date" means that a certificate is in the box.
I'm off to see the Wizard ...
Back from the land o' data.
I count 45 entries with inspection dates.
I count 27 entries that are marked "Wrong Sextant" (WS) or "Wrong Box" (WB) and that lack an inspection date.
There are 5 entries that are marked WS or WB among the 45 that have got inspection dates.
If you go look a the
list of sextants, you can change the sort order of the table by clicking on the small pair of triangles or arrowheads next to each field name. When you click on the "Inspection date" sort marker, all of the inspection date records will float to the top. So, too, with all the other fields.
Some of the sorting produces nonsense lower down the table. That's life with tables. If you get mixed up, just refresh the page and the table will start out again in its original state.
Time for coffee and pastry.
After breakfast -- counting coupGetting the right statistic out of the information in the table is a little tricky.
I made the table.
I'm not sure I'm counting things write.
Here is how things got complicated.
1. I wanted to keep track of
pairs of numbers.
2. After keeping track of numbers for a while, it became apparent that some sextants were in the wrong box.
3. Because of the mixups, we could get one pair of numbers from the sextant box and another pair of numbers from the sextant in the box. I put both numbers in the table.
4. "WS" indicates that it is an entry for numbers taken from the box, even though the box contains the "wrong sextant."
5. "WB" indicates that the entry is for a pair of sextant numbers engraved on the sextant, even though those numbers disagree with numbers on the box. This is how we know that the sextant is in the "wrong box."
6. Therefore, the same discrepancy may be recorded twice, I think: one entry for the box numbers, marked "WS," and one entry for the sextant numbers, marked "WB."
Checking the table now to see if there are the same number of WS and WB records.
...
Nope. 13 WS records and 19 WB.
Why? Because sometimes we could tell that there was a discrepancy between the Brandis numbers on the sextant and on the box but could not derive a pair of numbers from the information we had about the box.
Here is a weird record:
Brandis 4193
N.O. 4161
1919-04-02
Pictures show 4193 stamped on arc and inked on the box. N.O. 4161 also on arc, according to the text. Maker's number 5317 on the inspection certificate. So there is quite a discrepancy!
A theory to account for the apparent facts:
1. Brandis 5317 was put into the box for Brandis 4193 and sent to the N.O. for collimation.
2. The NO put a certificate for Brandis 5317 on the box and returned it to the sender.
3. Someone put Brandis 4193 back into its own box.
Cue sounds of wailing and gnashing of teeth.
OK. I'm off to do some more sorting and counting.
...
I've attached an ODS spreadsheet that makes it easier to resort the data and count things in the columns.
- 97 Naval Observatory numbers
- 7 of these entries lack a maker's number
- 90 pairs of numbers
- 31 of 97 entries are marked "WB" or "WS"
- 18 "WB"
- 13 "WS"
- 45 inspection dates recorded
- 4 of the 45 are marked "WS"
- 1 of the 45 is marked "WB"
- 26 of the 52 pairs of numbers that lack inspection dates are marked "WB" or "WS"
Should we say that each of the WB/WS entries is a separate event?
I'm inclined to say yes.
Every sextant put in a wrong box produced a correspondingly mismatched WB/WS pair.
If we start with 100 properly matched sextants and boxes, then determine that we have seen 15 mismatches, that (to me) implies another 15 that we haven't yet seen up close and personal.
31/97 is 32%.
Even boxes with collimation certificates do not always contain the right sextant. 5 out of 45 (11%) are mismatched.
For entries that lack collimation dates, the misplacement rate is 50%: 26/52.
It's possible that this big discrepancy is due to our inability to handle each sextant and box personally. It may be that the sellers or owners just didn't photograph the certificate. Because we weren't asking the question about a 1-to-1 ratio between instruments that had Naval Observatory numbers and boxes that had collimation certificates pasted inside them, we didn't keep track of whether the ABSENCE of an inspection date meant the ABSENCE of a certificate. I think some of the certificates had numbers on them but did not have the date when the adjustment was made--but I have been wrong in the past and could be wrong now.
Some of this information might be contained in the
thread about sextant numbers. I'm not ready to go back to read those 482 posts this morning. I have guests coming over for Mass and then for a quest to find cheap eats in the neighborhood.