ANIMAL PSYCHOLOGY IN FRANCE 453 



at right angles to each other, and on the return retraces them 

 successively in inverse order. 



This work of C.'s, as he expresses it, " undermines the theory 

 which attributes the return to the nest to muscular memory." 

 The sequence of movements on the return journey is often quite 

 different from that of the movements performed on the outward 

 course. When there is a change of direction, it cannot be attri- 

 buted to muscular memory; if, at such a moment, the ant is on 

 a movable support, and the support is turned through a certain 

 angle, the ant, disturbed, after having made various movements, 

 takes the same direction that she would have taken without the 

 interruption. In any case, there are two kinds of muscular 

 memory: memory of a sequence of movements, and memory of 

 the total work accomplished. According to C, the first does 

 not exist in ants, and if we admit the second, it is of a very crude 

 sort, and cannot explain the return to the nest. We have, in 

 fine, memory of one or of several successive directions, which 

 it seems to me cannot be explained except by associative memory 

 acting on the sensations furnished by certain landmarks. We 

 are thus brought back to the conception maintained by the 

 American investigator Turner. 



Many observations have been made on the habits of insects 

 and birds. As regards the latter, a certain number of studies 

 are reported in the Revue jrancaise d'ornithologie; the subjects 

 are the maternal instinct, nest -building, and migration. Among 

 the best observers of insects I may cite Ch. Ferton and E. 

 Roubaud. 



Ferton '*" among other things describes certain peculiarities of 

 nest-building (closure by partitions of parchment, building a 

 barricade of sand) and gives numerous details regarding the 

 prey of the Hymenoptera. We may note also certain curious 

 observations on the methods used by Pompilidae to paralyze 

 spiders, and on " attention in the return to the nest." 



The Synagris are solitary wasps belonging to the family of 

 the Eumenides and living exclusively in Africa. By the study 

 of three species, Roubaud '*> *" has been able to show instinct 

 evolving from the " once-for all " type of provisioning the nest 

 usual with the Eumenides, towards the continuous supplying of 

 food and daily care of offspring which is found in social wasps. 



