Autotetraploids 445 



endosperm nuclei is abnormal. The distribution of the chromo- 

 somes becomes very irregular, and the nuclei that result may 

 exhibit a considerable range of sizes and shapes. Cell walls fail 

 to form around the nuclei of the endosperm as they do about 

 forty-eight hours after fertilization in plants of barley. The 

 whole endosperm tissue sooner or later disintegrates, and the 

 seeds thereupon collapse. The embryo itself is normal, although 

 retarded, but eventually dies of starvation. The peculiar be- 

 havior of the endosperm in the hybrids is a secondary effect, 

 resulting from abnormal development of the antipodal cells. In 

 plants of the grass family, the antipodals enlarge greatly and 

 form a prominent tissue in the embryo sac that resembles secre- 

 tory tissue. They appear to secrete something necessary for the 

 normal development of the endosperm for about twenty-eight 

 hours after fertilization. In the hybrid, they fail to enlarge, 

 become dormant or almost so, and apparently fail to supply the 

 endosperm with substances necessary for its normal behavior. 

 The endosperm is abnormal and fails to supply the embryo with 

 its normal supply of food. 



Autotetraploids 



Autotetraploids are tetrasomic for all the chromosomes. Al- 

 though probably most tetraploids are allotetraploids, autotetra- 

 ploids have been found in nature and have been produced in the 

 experimental field. In general, they are slightly larger and more 

 robust than the corresponding diploids and are more adaptable 

 to different environm^ental conditions. They are not nearly so 

 sterile as autotriploids, but frequently they are somewhat more 

 sterile than the diploids from which they were derived. 



Since autotetraploids have their chromosomes in a tetrasomic 

 condition, it is to be expected that at least some of their con- 

 figurations at metaphase would be quadrivalents. Since, in zygo- 

 tene pairing, like parts normally pair and only two threads are 

 associated at any one place, a number of possibilities exist. If 

 all four threads of a tetrasomic group are tied in together at 

 zygotene, as in Fig. 126a, and if chiasmata form and terminalize, 

 the chromosomes will open out into a ring of four which re- 

 sembles in every way the ring of four produced by reciprocal 

 translocation. Such a ring is often found at first metaphase. If 

 chiasmata fail to form in one of the arms (Fig. 127^), the con- 



