1. It is already by one single reward (= one complete feeding act) that bees learn how to distinguish scentless sources from sources marked with flower scents and those from different ones in more than 90% of the trials. 2. In learning more than one scent successively the bees choose the first information preferably. The first scent becomes the determining factor. This is why bees keep to one sort of flower with great steadiness. 3. It is not only the position in the sequence of learning, but also the quality of the scents that effects the results of the tests. 4. The bees accept a scent after having been used to a source without a scent or with a different one without great difficulty. After 10 feeding acts at the second source and only one at the first the bees choose the second information in 95 resp. 86% of the test cases. 5. After receiving one reward with the first information and five with the second, the bees choose the first scent at 18% of the trials. The first information, however, is not forgotten but repressed. This can be shown by letting them choose between the first scent and an unknown third one. In this case the first information is chosen at 98.33% of the trials. 6. Social communication in the hive is sufficient for a new-comer to choose the communicated signal at the feeding place in 99.41% of the cases. 7. In learning only from their own experience (without any preceeding communication) one feeding act is sufficient for the bees to follow the learned scent signal in 96.74% of the cases. 8. A diagram of the efficiency in remembering scents taken over a prolonged period shows a drop but not a continuous one like an asymptote but with a periodical peak every 24 hours. The drop is not a function of the mere lapse of time but much more of the number of new and different bits of information that are stored in the memory during the interval between training and testing (thus explaining the difference between the results of Las Palmas and Frankfurt). 9. If the input of such additional information is stopped by locking the bees in the hive immediately after training, a grow in the efficiency of remembering can statistically be proved. © 1969 Springer-Verlag.
