818 POLYEMBKYONY AND THE DETERMINATION OF SEX. 



These facts, which we had crudely tried to attribute to occasional 

 parthenogenesis" (the caterpillars giving exclusively males being, 

 according to this supposition, those which had been stung by unfer- 

 tilized Encyrtus) we now explain in a far more rational manner. 



In man, genuine twins, inclosed in a single chorion, probably pro- 

 ceed from one egg. In spite of the different hypotheses which have 

 been proposed, particularly in recent years, on the subject of their 

 formation (Eosner, 1901) it is natural to admit that the twins 

 develop by the separation of the egg into two parts (spontaneous 

 blastotonw)- Now it has been determined that genuine twins are 

 always of the same sex. It is also known that, apart from some 

 extremely rare cases, there is identity of sex in double monstrosities. 

 The exceptions to this rule are probably exjjlained by the fact that 

 certain monstrosities form by the union of two eggs. 



Still another case presents itself in the mammals, and which, even 

 more than the preceding ones, seems comparable to those of Encyrtus 

 and Polygnotus. It is that of the armadillos {Dasypus or Tatusia). 

 This, in fact, does not involve an accidental case, but a phemomenon 

 of specific character; these animals bring into the world, according 

 to the species, a litter of 4 to 11 young, which are all and always of 

 the same sex. Now it Avas determined by vShering (1886)'' that all 

 the fetuses are enveloped in a common chorion and in consequence 

 belong to the type of true twins. Rosner (1901) had been able to 

 crudely explain the fact by the habitual presence of a number of 

 ovules within a single graafian follicle and had likewise concluded 

 that all the cases of monochorial budding pregnancy might be 

 explained in the same manner. But Cuenot (1903), reopening the 

 question, ascertained that in the species studied by Rosner {T. 

 novernc'mcta L.) the monovular follicles are five times more numerous 

 than the pluriovular ones. It is therefore impossible to admit that 

 the latter alone furnish fecundable eggs, and the author concludes 



a This observation is easy to repeat. Oue knows that tlie caterpillars of 

 Hyponomenta group their cocoons in nioi'e or less vohnninons clusters (nests) 

 attached to the branches. The chrysalids are to be found within the cocoons, 

 in June with the Hyponomenta of the spindle tree, in July with the Hypono- 

 menta of Prunus padus. The parasitized cocoons are immediately recognized 

 by their hard consistency ; it is easy to pull them apart and isolate each of them 

 in a vial (well dried), closed by a paper tied over the neck. After all have 

 issued the Encyrtus may be killed by putting a few drops of chloroform on the 

 paper. Then treating them with alcohol, oil of cloves, and Canada balsam, all 

 the Encyrtus proceeding from the same vial are mounted on a single slide and 

 one can then observe, under the microscope, the proportion of the two sexes. 



6 Shering examined two pregnant females of Tatusia hyhrida Desm. of 

 Paraguay ; each one of them contained eight fetuses, all masculine in both 

 cases, enveloped in a common chorion. 



