As you are in Dorset, I doubt you have cold summer nights, so you probably see the cumulative value shown in the new column does not change during the summer dates of May to September, so in effect for you (with a midnight rollover) incrementing does end after April.
Read
viewtopic.php?t=4736 for the background. The person requesting the feature says he is interested in knowing the value just for the winter half of the year (9 am 1 September to 9 am 1 April), but adds in a later post it could be simpler for coding purposes to count for whole year (the suggestion in that post says 1 September to 1 September). Steve Loft, however, coded it to start (when the counter is reset to zero) by default on
1 October.
The text in the Wiki was taken from the
"enhancement tracker" that Steve Loft hosted. The tracker was lost by Steve when he moved hosting, so although Freddie updated the link when he subsequently took over forum hosting that too fails.
Steve Loft
did predict that MX might store the figure.
Mark gave Chill Hours a makeover in MX (at 3.12.0 perhaps), the default start date (when the counter is reset to zero) now varies between 1 October and 1 April depending on what the MX user enters as Latitude says about their hemisphere of residence. The way it is calculated also changed in that makeover.
The dates I have used (since soon after Cumulus 1 introduced the feature on 2nd July 2011) are 9 am 1 September to 9 am 1 April, and originally I stored the values (taken from either the old this year web page, or from today.ini, as each winter ended) in a file. In 2014, I introduced a database to work with Cumulus 1, but my schema has always been different to the one later adopted by MX. Briefly, I store cumulative chill hours, and daily increment in a table with one row per day (for some columns day starts at 9am one day, some columns day starts at midnight, and other columns day start at next 9am, so close to UK Met Office practice; but all in same row), alongside columns for sensors I do have and other additional daily columns, but I exclude columns for sensors I don't have. I have rain related columns in a separate table that does not store rows for days when there is no rain. I also added another table storing the month-by-month chill hours increment alongside other derivatives taken from this month and monthly-all-time tags.