December 8, 1923] 



NATURE 



829 



into males, in the trout a much longer period, in 

 fact the utmost which the female parent will stand 

 -21 days' del^y — will only raise the proportion of 

 iiales to about 67 per cent. 



Working with the brown trout with shorter 

 4-14 days') periods of delay, I have obtained results 

 )f the same kind but not so marked.* 



It was thought that it would be interesting to try 

 !ie effects of delay outside the body. Accordingly, 

 11 December 13, 1922, all the ova of a large female 

 brown trout were stripped into a basin. One portion 

 (lot A) were fertilised immediately. The rest of the 

 eggs were left in the basin, in the small amount of 

 fluid which comes away with them, but without the 

 addition of any water, although in a moist atmosphere. 

 A second lot (B) was fertilised 2 days later, and a 

 third and final lot (C) after a total delay of 4 days. 

 The sperm of the same male was used for all three 

 fertilisations. 



The surviving young trout were preserved in 

 October 1923. All which had died after hatching 

 were ^.Iso preserved, and all but 4 of these could be 

 sexed. 



The results may be tabulated as follows : 



appear to be the only method by which it could 

 come about, but the reason for this remains obscure. 



.1 should like to take this opportunity of expressing 

 my thanks to IMr. Stevens, without whose interest 

 and help I could not have carried out the work. 

 The expenses were defraved out of a grant from the 

 Royal Society. ' Julian S. Huxley. 



New College, Oxford. 



PS. — I have just heard from Mr. Rowland Hazard, 

 the owner of one of the largest trout hatcheries in the 

 United States, that the sex-ratio of adult fish in his 

 experience varies considerably from year to year. This 

 year it has been about 58 per cent. $ $, irv 40,000 

 fish, but in most years the excess of males is less. 



None of the variation shown by these male percentages 

 is statistically significant. F"or example, the differ- 

 ence between the ^ percentages of A and B (totals 

 sexed) is 2-1, with a P.E. of ±3-39; and that 

 between those of A and C is 4-7 ±4-06. Even that 

 between those for " C, died after hatching" and 

 " C, killed Oct. 1923 " is only 21-5 + 7-84, and must 

 therefore be treated as an error of random sampling. 

 Thus the treatment has no effect upon the sex-ratio. 



The mortality rate, on the other hand, is markedly 

 affected by the treatment. This is especially notable 

 in the death-rate before hatching ; but even after 

 hatching, although the controls show the abnormally 

 large death-rate for the first ten months of 26-7 per 

 cent, (due to an exceptionally bad attack of fungus- 

 disease), that for the treated ova is more than twice 

 as great. Mrsic [loc. cit.) found that over-ripeness 

 within the body of the female exerted a deleterious 

 effect upon tlie young fish, as had been previously 

 shown for frogs by Witschi.' But in neither case 

 was the mortality nearly so great as in these experi- 

 ments. The increa.sed mortality was thus merely 

 something incidental to any abnormal treatment ; 

 but the two treatments, of delay inside and delay 

 outside the body, exert quite different effects in other 

 respects upon the unfertilised ova. 



The sex-ratio of adult (2- and 3-year old) trout, 

 both brown and rainbow, reared in captivity is, 

 I am informed by Mr. Stevens, Manager of the 

 Midlands Fishery at Nailsworth, close to 66 per cent. 

 (5 cJ in his experience. What brings about this 

 marked increase in the number of males is not known 

 (in all recorded experiments the sex-ratio of young 

 nsh 6-10 months old is close to 1:1). Differential 

 elimination of females after attaining maturity would 



» Huxlry, 1923, Science, 58. 291. 



• Witschi, 1922, Biol, /.cnlralhl. 42. 97. 



NO. 2823, VOL. I 12] 



Is the Pentose of the Nucleotides formed under 

 the Action of Insulin ? 



We have read with interest the letter from Mr. 

 C. Berkeley in Nature of November 17, p. 724. 

 Referring to the substance found in animal tissues 

 after insulin, which gives the a-naphthol reaction, 

 but is without reducing action on copper salts, he 

 suggests that our failure to find 

 reducing power after hydrolysis by 

 acids may be due to a pentose 

 constituent going over to furfurol 

 and being lost by volatilisation. 



There are reasons which make it 

 unlikely that this is the case. We 

 find that on boiling a solution of the 

 substance in the presence of 8 per 

 cent, hydrochloric acid for three 

 hours, there is no measurable change 

 in the optical properties of the solu- 

 tion. This does not exclude the 

 possibility of loss of pentose sugar 

 owing to conversion to furfurol, since the optical 

 properties of the residue may have been altered, 

 with the result that no change would be observed. 

 On the other hand, the substance gives little indica- 

 tion that it contains a pentose. The modification 

 of the a-naphthol test in which strong hydrochloric 

 acid is used in place of sulphuric would be likely to 

 indicate the presence of pentoses, by a rapid appear- 

 ance of a purple colour. 



Using this test, a positive reaction is obtained only 

 shortly after the sugar has been extracted from the 

 tissue, and then only faintly. The substance is 

 dried with difficulty at room temperature in vacuo 

 over sulphuric acid. At the end of this period the 

 a-naphthol reaction (using sulphuric acid) is given 

 with diminished intensity, and in the course of weeks 

 it may disappear entirely. The a-naphthol test is 

 extremely sensitive, and since many proteins and 

 their derivatives give a positive reaction, it is perhaps 

 necessary to emphasise that the intensity of the 

 colour given by the substance in question, as well as 

 the method of extraction employed, point to the 

 substance being of a true carbohydrate nature. 

 Indications have been obtained that the substance 

 is present in normal tissues. 



Dudley and Marrian [Biochem. Jour. 17, p. 435, 

 1923) have shown that glycogen disappears almost 

 entirely from the liver and muscles after insulin 

 convulsions. The fact that the blood sugar of 

 animals in convulsions may be restored to the normal 

 level, and the animals be recovered, by injection of 

 such substances as adrenalin or pituitrin, without 

 injection of glucose, suggests that the carbohydrate 

 reserves of the body have been converted into some 

 form other than glycogen. That the Islets of Langer- 

 hans tissue contain unusually large amounts of 

 pentose compounds is of great intorrst ; but this 



Z 2 



