STUDY SUBJECTS WERE 80 PATRONS -- INCLUDING 23 NONSTUDENTS -- OF A UNIVERSITY LIBRARY. THE AVERAGE AGE OF THE SUBJECTS WAS 25. THE SUBJECTS READ FICTITIOUS ACCOUNTS OF THE MURDER OF A YOUNG CHILD DURING A RESTAURANT HOLDUP. HALF OF THE SUBJECTS WERE LED TO BELIEVE THAT THE DEFENDANT'S ACCOMPLICE HAD ELUDED CAPTURE; THE OTHER HALF WERE TOLD THE ACCOMPLICE HAD BEEN APPREHENDED BY POLICE. AS A CROSS-DIMENSION, HALF OF THE SUBJECTS READ AN ACCOUNT DESCRIBING THE DEFENDANT AS PRIMARILY RESPONSIBLE FOR THE CRIME, WHILE THE OTHER HALF WERE TOLD THAT THE ACCOMPLICE WAS MORE RESPONSIBLE. SUBJECTS WERE ASKED TO ASSIGN AN APPROPRIATE NUMBER OF YEARS OF PUNISHMENT TO THE DEFENDANT AND TO DESCRIBE THEIR RESPONSES TO THE CRIME. SUBJECTS DID NOT DEAL WITH THE DEFENDANT MORE LENIENTLY WHEN THE ACCOMPLICE ESCAPED. SUBJECTS WHO REPORTED FEELING RELATIVELY LESS ANGRY ABOUT THE CRIME TENDED TO ASSIGN LESS PUNISHMENT WHEN THE ACCOMPLICE ESCAPED. CONFLICTING FINDINGS FROM EARLIER STUDIES ARE NOTED, AS IS THE NEED TO INVESTIGATE FURTHER THE ROLE OF JURORS' EMOTIONAL REACTIONS IN DECISIONMAKING. TABULAR DATA AND A LIST OF REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)