The lenses have indeed changed a little over the years, and you should not switch lenses across systems.
Each system should have one infant lens for use with the 940nm (infant) illuminator and another, "default" lens that should be used with the 890nm (default) illuminator. The wavelength of the illuminator should be marked on a sticker on the side of the illuminator.
Infants up to 6-9 months of age have slightly different irises than the rest of us do, and the use of the 940nm illuminator provides an image with better contrast (between pupil and iris) than the use of 890nm illuminator does. For each system, one lens is optimized for use with the 890nm illuminator and the other is optimized for use with the 940nm illuminator.
So, the decision making process is like this:
1) First, decide which type of illuminator you should use based on the age range of the participants. If it will include participants less than 9 months of age you probably want to use the 940nm illuminator (this will also work on older participants -- it may have slightly higher noise, but should still be able to track them). If your participants will all be older than 9 months then you can use the 890nm illuminator.
2) Then, based on that choice of illuminator, use the infant (940) or default (890) lens.
It sounds like you have already identified which lenses are infant vs non infant, but if you would like me to double check (maybe for the Duke system, which it sounds like didn't have stickers on it), then could you please send me pictures of the two lenses for that same system? That will help us to distinguish which should go with the 940 and which with the 890.
