independently, but muscle activity of the left and right

zygomatic regions can be stimulated separately.

 During affective imagery, positive emotional states

 have been associated with decreased corrugator and increased

 zygomatic activity. Conversely, negative emotional states

 have been associated with increased corrugator activity and

 decreased zygomatic activity (Schwartz et al., 1976a,

 1976b). Also, when verbal report of emotions has been

 obtained, corrugator activity positively correlates with

 unpleasant emotions and negatively correlates with pleasant

 emotions. The opposite pattern has been found for zygomatic

 activity (Brown & Schwartz, 1980; McCanne & Anderson, 1987;

 Slomine and Greene, 1993). Similar results have been

 reported from other investigators using self-referent

 statements designed to induce either elation or depression

 (Sirota, Schwartz, & Kristeller, 1987), and affective slides

 (Cacioppo, Petty, Lasch, and Kim, 1986). Additionally, an

 interview technique was employed to elicit and investigate

 naturally occurring emotional states (Cacioppo, Martzke,

 Petty and Tassinary, 1988). Replicating previous findings,

 elevations in corrugator EMG were related to lower positive

 emotion ratings and higher negative emotional ratings.

 In sum, the above studies attest to the importance of

the covert activity of the corrugator supercilli and

zygomatic major muscles as indexes of emotion.

Specifically, EMG activity of the corrugator supercilli has