NavList:

The topic is quite large, ranging from chronometer history, chronometer practices, how to rate the chronometer, how many to have on-board, etc, etc, etc.  Most CN books do cover these topics.

Here is a very brief overview, just to get you started.  I will assume we are talking about mechanical chronometers.

Three chronometers are recommended on board, as if one goes wrong or bad, it will be evident from the other two, which remain in agreement.

The chronometers must be protected from environmental changes, consequently, they are often carried in nesting boxes, with glass lids so they can be viewed.

Chronometers are wound each day, at the same time, in the same number of rotations, so as best to preserve the rate of timekeeping.

Chronometers are kept in a cabin.  Time is transferred to the observation post via a hack (or deck) watch.

The chronometer face value should be recorded for reference as you go, making the deranged chronometer obvious.

The face time of the chronometer is NEVER changed.  The chronometer error (CE) represents the delta between the face value and true (GMT, usually) time at the instant of rating.  It the proper, true time is required, simply add the CE to the face value (but take care of CR, which follows).  The current CE replaces the previous CE.  That is, whenever you determine a new chronometer error, it represents the absolute delta.  The older CE is discarded.

Chronometers rarely run at precisely the same rate as time itself.  That is, the chronometer runs faster or slower than actual time.  This is the chronometer rate (CR).  A CR=0 means that your chronometer keeps perfect time.  A CR=slow 7 seconds means that the chronometer is losing 7 seconds PER DAY.  To determine true time, we also must therefore add the CR multiplied by the number of days since we rated the chronometer.  Fractional days should be considered for the utmost in accuracy.

To review

1) Your chronometer has an arbitrary face value.  

2) You add the CE to get true time at the instant the chronometer was rated.  At thst instant, the number of days since rating is zero, so the CR term drops out.

3) You multiply the number of days (and fractional days if desired) by the CR, to get the true time at the instant of observation and add it to the vslye found in step 2.

How to determine the CR and CE is an interesting topic.  CE can be determined, like Worsley did, by inverting the observation.  You are at a known location (Cape Belsham) and observe an altitude (jargon alert!) and solve for time!

The CR requires that you solve two CE, days apart, to determine how much you gain or lose over a period.  Thus you get CR.

This topic is incredibly broad.  Each statement above has volumes of information, practice and tradition associated with it.  

Brad