Harry Harlow set up one of the nation's best-equipped primate laboratories at the University of Wisconsin. His famous study focused on contact comfort in babyhood. He did an experiment with a rhesus monkey to see which would be more powerful, contact comfort or food.
In this experiment, he separated the baby monkeys from their mothers only a few hours after birth. He provided the babies with a choice between two fake mothers, one made out of uncomfortable wire but with nourishment, a bottle of milk, and the other with a layer of comfortable terry cloth.
Harlow found that baby monkeys preferred the terrycloth mother, spending as little time as possible on the wire frame mother. This overturned the previous assumptions that the primary basis for the attachment bond is the nourishment supplied by mothers. This study showed that contact comfort was more important than food, except when the baby monkey was actually eating. And when Harlow exposed monkeys to a scary stimulus, like a toy robot playing a drum, they were more likely to run to the terry cloth mother and cling to her for reassurance.
I found this study interesting because it seemed to prove a large phenomenon without any questioning. Since Harlow did not use real Rhesus monkey mothers in his experiment, one could potentially also conclude that a soft and cuddly mannequin would suffice just as well for a human baby as its mother. It seems like another study could be done with alive monkey mothers to prove this contact comfort theory further...