These 4LR44 batteries have not lived up to my initial tests.
I thought I had done due diligence by endurance testing the little six volt batteries powering the MK111 timer. I was using two 4LR44 batteries which fit nicely in a holder for an AA battery. One pair of these had run the MK111 timer for two test flights. The first flight lasted 27 minutes and captured 110 NIR/visible pairs of photos. The same batteries then flew for two more hours and captured 265 additional pairs of photos. The timer was sending a pulse to the cameras every eight seconds.
So I was disappointed when I reeled the kite in last week after an hour and fifteen minutes of walking it along a half mile transect of tundra study plots. I thought I was capturing continuous coverage of a five acre area and was imagining the awesome Hypr3D model I would make from the images. But I learned an hour later that the fresh timer batteries had died after only 13 minutes. I put fresh batteries in and relaunched, but the batteries were dead again after 19 minutes. The pair that lasted 13 minutes was shooting every eight seconds (sending a pulse every four seconds), and captured 119 pairs of photos. The pair that lasted 19 minutes was shooting every 16 seconds (sending a pulse every eight seconds) and captured 76 pairs.
I am not sure why the test flights saw such better performance from the batteries. It was about 20 degrees F warmer for the tests (85 vs. 65) so that could explain at least some of the difference. These 4LR44 batteries are apparently not up to the task of running the MK111 timer reliably for typical mapping flights of 30 to 90 minutes. My plan is to modify the MK111 timer to operate without the relay (which requires most of the power) following the suggestion of Randy Sargent in the comments on this note.
Fortunately I had a backup timer device (AuRiCo), but it started to rain soon after I launched it and I never did get full coverage of that transect.