Apeil 30, 1909] 



SCIENCE 



705 



chromosome, but rather the absence from the 

 G-chromosome of a particular pigment-pro- 

 ducing factor. I would therefore regard it 

 as a more plausible guess that a T-element is 

 present in both sexes, and that both have the 

 same number of chromosomes, the female 

 zygote formula being XY and the male TT, 

 as the facts in the sea-urchin suggest. The 

 female heterozygote thus becomes GLXT, the 

 male GLYY, and the homozygous male GGYY 

 or LLYY. All the facts are then consistently 

 accounted for by the single assumption that 

 G, while acting as the synaptic mate of L, 

 always undergoes also a secondary coupling 

 with Y. 



Did such secondary coupling not take place 

 the female GLXY would give rise to the bival- 

 ents G/L and Y/X, producing the four classes 

 of gametes GX, GY, LX and LY. If, however, 

 in addition to the primary synaptic coupling of 

 X and Y, G also couijles secondarily with Y, 

 the result should be a quadrivalent element, 

 which might have either the tetrad grouping 



GY 

 LX 



or the linear grouping 

 G 



L 



giving in either case the two classes of gametes 

 GY and LX. In the males, GLYY or GGYY, 

 the gametes will of course be GY, LY or 

 GY, GY respectively. This gives a series of 

 formulas identical with those of Bateson and 

 Doncaster, as recast by Castle, if Y be every- 

 where inserted in its proper place, as follows: 



Parents Constitution Gametes Oflspring 



(1) I/ffic*. ? LLXY LX,LY Q'LX.Y — gross. <i. 

 Gross. S GGYY GY, GY GLYY = (/ross. <? 



(2) Hei.2 GLXY GY, LX GGYY = gross, c? 

 Bet.^ GLYY GY,LY GLXY = gross. ? 



(3) Lao*. 2 LLXY LX, LY GLXY = gross.? 

 Bet.^ GLYY GY,LY GLYY = gross, c? 



LLXY = lact. 5 

 LLYY = lact. g 



(4) ffei. $ GLXY GY, LX GLYY = gross, c? 

 Lact.^ LLYY LY, LY LLXY = Zac*. $ 



This adds nothing in principle to Castle's 

 suggestions, but seems more in accordance 

 with eytological expectation. 



Such a mode of coupling may seem very 

 improbable ; but I wish to point out that there 

 are at least some approximate analogies to it 

 in eytological facts known in other animals. 

 Several different types of multiple elements, 

 formed by definite chromosome-couplings, are 

 now known. An example is given by Meta- 

 podiics (which I have recently described in 

 detail). In individuals having "supernu- 

 merary " chromosomes these regularly couple 

 with the idiochromosome-bivalent in the sec- 

 ond division to form triad, tetrad, pentad and 

 even hexad complexes; and the components 

 are often arranged in linear series. I have 

 recently obtained an individual of M. femor- 

 atus which differs from all other individuals 

 of the species thus far examined in possessing 

 a single odd or accessory chromosome, while 

 the missing small idiochromosome is replaced 

 by a third " »i-chromosome." The latter does 

 not, as might have been expected, play the 

 part of a synaptic mate to the odd chromo- 

 some, but shows throughout the spermato- 

 genesis the characteristic behavior of its own 

 kind. In the first division it is always coupled 

 with the two other m-chromosomes to form a 

 triad element, the three components almost 

 always forming a linear series. Again, in 

 Thyania there are three sex-chromosomes (the 

 Y-element and two components of the X- 

 element) which divide separately in the first 

 division but are always coupled in the second 

 to form a linear triad series. In the reduvi- 

 oids, as Payne has recently shown, the sex- 

 chromosomes form in the second division dyad, 

 triad or tetrad groups; in Gelastocoris they 

 form a pentad complex; and in each case the 

 components show a definite arrangement and 

 mode of distribution. 



A closer approximation to the secondary 

 coupling suggested in Abraxas is given by the 

 observation of Sinety on Leptynia, (one of the 

 Phasmidse), and especially by the discoveries 

 of McOlung in Sesperotettix and some other 

 Acrididffi, that the X-element (accessory ehro- 

 mosorae) is in these cases regularly coupled 

 in the maturation-divisions with one of the 



