After looking at Web images and viewing plate XVIII in Dillons, I concur that these little guys are genus Stenelmis, family Elmidae. Number of antennal segments differentiates genera, with Stenelmis having 11 segments while the other genus has only 7. I'll try to get a close-up in which segments can be counted, but general shape and markings leaves no doubt that it belongs in Stenelmis.

This is an aquatic family, with adults and larvae often found clinging to rocks in running streams.

Jim: Can't see as well as I would wish, but am pretty sure that this is Stenelmis crenata - by FAR the most common Stenelmis in New Hampshire.

If you care to have another look, perhaps the two close-ups I've added will help cinch your ID. These are of a dead specimen.

Unfortunately as with most all the keys in Dillon, it is extremely misleading and nearly useless at the genus and species level. There are at least 6 or 7 genera of elmids in the Northeast alone...with several times that number in North America. Many of them are common. The # of antennal segments will not be much help and is not a key character, except for perhaps excluding Macronychus, which is a wierd monotypic genus anyway with that lives on submerged wood and root masses. (But yes, in this case, it is Stenelmis...). One of the best resources for the ID of aquatic Dryopoidea is a bit outdated, but still excellent:

1972. Aquatic dryopoid beetles (Coleoptera) of the United States. Biota of Freshwater Ecosystems Identification Manual No. 6. Water Pollution Conference Series, United States Environmental Protection Agency, Washington, District of Columbia.

The key in American Beetles is good, but not easy for the non-specialist.

Should I leave it at just Elmidae then?

I think it is safely Stenelmis