PLASTIDS AND CHONDRIOSOMES 119 



the grasshopper, Dissosteira Carolina, Chambers (1915) finds that the 

 chondriosomal material forms a granular network surrounding the nucleus 

 during the resting stages and the mitotic figure during division. During 

 the later phases of mitosis the strands and granules of this network 

 lengthen into delicate filaments between the two daughter chromosome 

 groups, and finally separate into two granular masses which gradually 

 invest the daughter nuclei. 



In the mole cricket, Gryllotalpa borealis, the distribution of the chon- 

 driosomes to the daughter cells is accomplished with even greater defin- 

 iteness. According to Payne (1916) they become thread-like and break 

 near the middle, the halves passing to the daughter cells. Voi'nov (1916) 

 states that the "mitochondria" in the spermatocyte of G. vulgaris fuse to 

 form a thread which then segments into a number (70 or more) "chondrio- 

 somes." These are arranged on the spindle along with the chromo- 

 somes, which they may closely resemble, and divide to form daughter 

 bodies at both maturation divisions, so that they are equally distributed 

 to the four resulting spermatozoa (Fig. 44, A-C), 



In certain scorpions also the chondriosomal material is distributed 

 with surprising precision. In a species from Arizona (Wilson 1916) this 

 material in the spermatocyte takes the form of a single ring-shaped body. 

 This ring divides accurately, much like a chromosome, at both maturation 

 divisions, each of the four spermatids, and hence each spermatozoon of 

 the tetrad, receives a quarter of its substance. In a California species 

 (Wilson) there is no ring formed, but instead about 24 hollow spherical 

 bodies. At the two maturation divisions these show no evidence of 

 division, but are passively separated into four approximately equal groups, 

 each spermatid receiving six (occasionally five or seven). A European 

 species described by Sokolow (1913) agrees essentially with this. 



Function. Our knowledge of chondriosomes is yet too incomplete to 

 warrant any categorical statements regarding their functions, but a 

 number of opinions have been expressed, some of them based upon ob- 

 servational evidence and others upon conjecture. Certain of the more 

 prominent opinions may here be reviewed. 



It was in 1897 that Benda suggested that chondriosomes might be 

 distinct cell organs with a special function. In a series of papers which 

 began to appear ten years later Meves (1907 etc.) put forth and empha- 

 sized the theory that they play an important role in heredity that they 

 carry the hereditary characters of the cytoplasm. Evidence supporting 

 this view was seen by Meves and Benda in certain experiments of God- 

 lewski which seemed to show that the appearance of certain hereditary 

 characters is dependent upon something present in the cytoplasm rather 

 than in the nucleus. (See Chapter XIV.) This theory has had the 

 support of a number of investigators, among whom are the botanists 

 Cavers (1914) and Mottier (1916). Vo'inov (1916) also believes that the 



