No. 1, February, 1921J 



GENETICS 



39 



tumor character in only five per cent of the flies shows other genes also involved. These are 

 under investigation. 780 larvae show from one to three tumors situated in 6, 8, 9, 10, 11, 12, 

 13 or 14th segments. Metastases may or may not be present. Tumor may also be in head or 

 may replace wing or other appendages. Tumor cells are rounded or polygonal and contain 

 pigment. Pigment increases with age. Cells giving rise to tumor are originally hypodermal. 

 Tumor in adult fly is permeated with black pigment. When tumor develops in abdomen, 

 no shortening of life in fly results. Tumor was inoculated in 40 larvae; only 2 survived; both 

 grew tumor and were sterile females. — C. C. Little. 



243. Tackholm, G. On the cytology of the genus Rosa. A preliminary note. Svensk. 

 Bot. Tidskr. 14: 300-311. 3 Jig. 1920.— Very brief summary of author's chief results. About 

 300 bushes, representing approximately 230 different forms, are investigated. The funda- 

 mental haploid chromosome number in Rosa is 7. There are two very different groups of 

 species: (1) the very polymorphous conmo-section (in the widest sense), and (2) all other 

 sections. The latter group have only gemini (7, 14, or 21) in the diakinesis; the former pos- 

 sess both paired chromosomes and single ones. The following table gives the observed 

 numbers: 



The roses of types l-3^have only sexual reproduction. Some of this species have varieties 

 with different chromosome numbers (haploid number 7 and 14; 14 and 21). Triploid roses 

 (type 4) may be hybrids between types 1 and 2. Type 8 may be interpreted as the result 

 of a cross between types 2 and 3. The reduction division in the anthers is quite in accord- 

 ance with the Drosem-hybrid scheme of Rosenberg. The immense number of forms belong- 

 ing to the section Caninae (in the widest sense) are distributed in the types 5-7. The 

 reduction divisions of the Caninae in the anthers and the ovules are carried out in a quite 

 different manner. In the pollen mother cells, the gemini first come to the equatorial plate. 

 Later the univalent chromosomes which are scattered irregularly on the spindle are also arranged 

 in the same plate surrounding the gemini. The partners of the bivalents first pass to the 

 poles. The singles lag a little, divide, and then the halves pass to the poles. The bivalent 

 and univalent chromosomes consequently have different anaphase stages. Not all the 

 halves of the singles may be included in the daughter nuclei. By the second division the de- 

 scendants of the bivalents are regularly distributed and included, without or with a few of 

 the other chromosomes in the four ordinary microspores. There even arise a great num- 

 ber of small microspores from each mother cell, containing only descendants from the uni- 

 valent chromosomes. Some of the microspores — probably those containing descendants of 

 the bivalents— develop into pollen grains, are able to germinate, and must be supposed to be 

 capable of fertilization (hybrids between roses in the section Caninae are known). In the 

 embryo-sac mother cells we have ordinarily the two equatorial plates — the bivalents next 

 to the chalaza, the singles next to the micropyle. In the anaphase the partners of the gemini 

 are separated; but in about four-fifths of the studied cases, all the singles pass undivided to 

 the micropylar pole. In the second division all the chromosomes are divided. Consequently, 

 the tetrad has two large nuclei and two smaller ones, with only 7 chromosomes. The former 



