35 6 JOHN BELLIXG. 



temperature in February and March, when the greenhouse fires 

 were banked. 



The preparations of young microspores and pollen-mother-cells 

 were fixed and stained in iron-acetocarmine (Belling, 1921). A 

 water-immersion objective was used to examine the chromosomes. 

 It may be noted that an apochromatic water-immersion of 1.25 

 aperture will allow a true condenser aperture of 1.2, if the source 

 of light is diaphragmed to fit the field of view. The working 

 aperture of the combination is about 1.2. On the other hand, 

 the working aperture of the 1.4 N.A. apochromatic oil-immersion, 

 used with a dry Abbe uncorrected condenser on objects in iron- 

 acetocarmine, is only 0.9 or less. (Compare Belling, 1923.) 



Three hypotheses have been made as to the rings seen at the 

 metaphase in certain Liliaceae and other genera of flowering plants, 

 (i) That these rings result from the mere twisting of the con- 

 stituent chromosomes of the bivalents around one another, and 

 that these become untwisted and longitudinally split at the 

 anaphase. This hypothesis is, it would seem, negatived by the 

 fact that, with the best visibility in the microscope, it is usually 

 impossible to say which chromosome is above and which below 

 at the junctions. (2) That the rings are caused by alternate 

 openings between the homologues and between their constituent 

 chromatids, so that adjacent rings are always at right angles. 

 This hypothesis does not appear to agree well with the state of 

 affairs in the trivalents of the triploid hyacinths, where 3 chro- 

 mosomes are concerned (Belling, 19250). (3) That between 

 each two adjacent rings there is a chiasma in Janssens' sense 

 (or a similar interlacing of strands) where segmental exchange 

 may have taken place, by the fracture and subsequent reciprocal 

 junction of two of the four strands. This assumption seems 

 most suitable for a working hypothesis, since crossing-over (and 

 hence probably segmental interchange) has been shown to occur 

 in those flowering plants which have been investigated ad hoc. 

 However, in the present paper only the following points are 

 regarded. 



1. Whether in horizontal rings or Y's one chromatid passes 

 up and one down from each lateral half of the ring or V. 



2. \Yhether these chromatids in separating show signs of inter- 



