Another test was done with slightly older infants at bedtime. In some groups the room was silent; in others recorded lullabies were played. In others a ticking metronome was operating at the heart-beat speed of 72 beats per minute. In still others the heart-beat recording itself was played. It was then checked to see which groups fell asleep more quickly. The heart-beat group dropped off in half the time it took for any of the other groups. This not only clinches the idea that the sound of the heart beating is a powerfully calming stimulus, but it also shows that the response is a highly specific one. The metronome imitation will not do—at least, not for young infants. So it seems fairly certain that this is the explanation of the mother's left-side approach to baby-holding. It is interesting that when 466 Madonna and child paintings (dating back over several hundred years) were analyzed for this feature, 373 of them showed the baby on the left breast. Here again the figure was at the 80 per cent level. This contrasts with observations of females carrying parcels, where it was found that 50 per cent carried them on the left and 50 per cent on the right. What other possible results could this heart-beat imprinting have? It may, for example, explain why we insist on locating feelings of love in the heart rather than the head. As the song says: 'You gotta have a heart!' It may also explain why mothers rock their babies to lull them to sleep. The rocking motion is carried on at about the same speed as the heart-beat, and once again it probably 'reminds' the infants of the rhythmic sensations they became so familiar with inside the womb, as the great heart of the mother pumped and thumped away above them.