There is an almanac in Manchu at the Bibliotheque Nationale, which lists (among many other interesting things) the time at which the sun was expected to rise and set at different locations in Mongolian and Manchu areas on calendrically important dates in 1769. The tables in this almanac give us examples of how to state the time in Classical Manchu down to the minute.
The time of day was expressed in three parts: erin, kemu and fuwen. There were twelve erin in one day, so a single erin was 120 minutes long, and they were named after the cyclical animals. The day started at 1:00 AM with the hour of the ox.
There were eight kemu in one erin, so each kemu was fifteen minutes long, and they were divided into two groups. The first group were the ujui kemu and the second group were the tob kemu. Of each group, the first kemu was called uju ("head"), and the other three were numbered one through three. This is interesting in part because the word ujui is usually synonymous with emuci (both usually meaning "first"), but here they have different meanings.
There were fifteen fuwen in one kemu, so the fuwen is identical to one minute, and they were simply numbered one through fourteen, with the first fuwen left unnamed.
So, if you wanted to say 2:57 PM, you would start with the erin, which would be honin (starting at 1:00 PM). Within that erin you would be in the second half (tob), and three full kemu will have already passed, putting you in the tob ilaci kemu (2:45 PM - 2:59 PM). Within that kemu you would be in the twelfth fuwen.
As with dates and places in Manchu, you always start big and go small. So, if you wanted to express 2:57 pm on the 8th of February, 1769, you would write it as follows:
