More from the same magazine:
In 1932, a biologist named Dr. Lee R. Dice noticed an unusual behaviour in a common house mouse that he caught in Michigan - he saw it singing. As Dice reported in the Journal of Mammology, the mouse's songs were "bird-like," erratic, easily localized, and usually heralding an rexcited state.
Documents from the same period reported similar behaviour not just in Dice's mouse(Mus Musculus, but in other mouse species, including harvest mice (for example,Reithrodontomys fulvescens),
Asiatic strains of Mus, and some species of the deer mouse, Peromyscus. The songs of all these mice had a defined harmonic structure with rates of between two and six notes per second.
Since then, biologists working on a number of Murine (the common house mouse and its relatives) and deer mice have shown that regular vocal communication is more common in mice than once thought. Using specialized equipment, biologists have for some time routinely observed mice whistling and reptitively calling at ultrasonic frequencies, inaudible to the human ear. Ultrasonic signalling is less likely to be heard by potential predators; although many animals can hear higher frequencies than humans can, the range of optimal hearing in many predators is much more restricted.
What is surprising, though, is that singing that is audible to the human ear, as in Dice's mouse, is common in a variety of mice, which regularly and preferenyially vocalize in this way. For a small mammal under high risk of predation, this audible signalling provokes some perplexing questions about how vocal behaviour evolved in mice...