208 Cytoplasm as Seat of Genetic Properties 



in Lynuintria, though here a recessive action is not modified but is 

 completely suppressed in one cytoplasm, strangely enough, its own. 



Another set of facts may point in the same direction. It is gener- 

 ally assumed that the sperm does not carry cytoplasm into the egg. 

 But both sperm head and middle piece contain mitochondrial prod- 

 ucts in the form of spiral fibrils and derivatives of the mitochondrial 

 body. In a little-known paper, Held (1916), a very good histologist, 

 showed that in Ascaris (where the sperm contains a considerable 

 amount of cytoplasm) egg and sperm mitochondria may be stained 

 differentially. He could trace the sperm mitochondria into the egg 

 and tlieir dispersal among the mitochondria of the egg. (Held was 

 a follower of Meves, who thought that the mitochondria are the real 

 genie material, not the chromosomes.) We might conclude that simi- 

 lar behavior of the mitochondria occurs in fertilization with filiform 

 sperm. An additional fact is known for Lepidoptera. Here normal 

 polyspermy occurs and each sperm head transforms into a nucleus 

 surrounded by a bit of specifically staining cytoplasm (see photos 

 in Goldschmidt and Katsuki, 1928f?). It is not proved that this 

 cytoplasm is derived from the sperm. But if it is, the 6-10 sperm heads 

 in an egg might contribute a rather large amount of paternal mito- 

 chondria. 



If this is true and if the cytoplasmic effect of shifting genically 

 controlled reactions is actually based upon mitochondria, we should 

 expect to find paternal cytoplasmic effects also (excluding here such 

 processes as transfer of chloroplasts by the pollen, to be mentioned 

 below). Actually such effects were recorded for the same material 

 as before, the pigment in Lymantria caterpillars. I had noticed early 

 (but had not dared to report on it for years in view of the general 

 attitude at that time**) an effect which could be explained only in 

 terms of an action of the sperm cytoplasm. Only after it was found 

 repeatedly did I present the data (Goldschmidt, 1924). Double re- 

 ciprocal F2 crosses, for example (Axa) X (Axa) versus (A X a) 

 X (a X A), are identical in regard to Mendelizing genes and egg 

 cytoplasm (always A in the example). But the presumed sperm cyto- 

 plasm is different (derived from the paternal grandmother), namely, 

 a in the second cross. Many such double reciprocal crosses in both 

 directions were checked for the pigmentation character reported 

 above, and in many instances differences in the curves of pigmenta- 

 tion were found between the two double reciprocal crosses, which 



" I had reported on it in an unpublished talk at Bateson's invitation in the 

 John Innes Institute. 



