Maech 18, 1921] 



SCIENCE 



263 



itr of the ova would be properly inhibited 

 but occasionally an ovum would fail to re- 

 spond because of some peculiarity in its 

 organization. Suppose that such peculiarities 

 are due to local factors appearrug with a 

 frequency of 1/n. Then the chance that two 

 such independent local factors will act at the 

 same time and thereby cause the simultaneous 

 discharge of two supernumerary ova is 1/n-. 



2. Monozygotic Twins and Triplets. — Nor- 

 mally a single embryonic area appears in the 

 blastodermic vesicle and through some coordi- 

 nating mechanism inhibits the formation of 

 additional embryonic areas. Suppose that 

 once in n cases a cell or group of cells ac- 

 quires physiological indei)endence as the re- 

 sult of an intrinsic factor and forms a second 

 embryonic area. The chance that two such 

 cells or groups of cells will arise at the same 

 time is once in n squared if it is supposed as 

 in the previous case that the two events are 

 independent of each other. 



In order that the stated numerical relations 

 may ensue, the important consideration in 

 ■either mode of origin of multiple births is the 

 independence of the two events which give rise 

 to triplets. If, in the fluctuations of the gen- 

 eral physiological state of the mother, the con- 

 dition is sometimes such as to result in twins 

 and sometimes in triplets, it is hard to see why 

 the " square " relation should exist. For in- 

 stance if it is postulated that additional ova 

 are stimulaited to complete the maturation 

 process as a result of an unusual amount of 

 an internal secretion and that the number of 

 extra ova depends on the quantity of the secre- 

 tion there is no reason for expecting the ob- 

 served relation between one extra and two ex- 

 tra embryos. This difficulty seems to ajpply to 

 all general agents that may be postulated as 

 acting upon the ovaries as a whole in the 

 cases of multiple ovulation or upon the devel- 

 oping enibryo as a whole in the case of mono- 

 zygotic twins and triplets. If, however, each 

 supernumerary ovum is due to an independent 

 local action and such local actions have a cer- 

 tain average frequency the coincidence of two 

 such a'ctions would give the observed numer- 

 ical relation of triplets to twins. 



If the explanation as stated applies to the 

 relation between triplet and twin births it is 

 to be expected that it will apply to quadruplets 

 as well. In that case the expected number of 

 quadruplets is one in n^. Unfortunately the 

 numbers are too small for a reliable conclu- 

 sion. In the largest available collection of 

 data, the one mentioned above, there are 36 

 quadruplets in 13,360,557 births or one in 

 (71.9)^ which is somewhat greater than the 

 expected number, one in (89.1)^. 



As in other statistical relations the biolog- 

 ical significance in the present instance can 

 not he proved directly from the mass of data. 

 When one considers the vicissitudes of fertili- 

 zation, ithe chances of death ,of individual em- 

 bryos, the demonstrated influence of the sper- 

 matozoon in certain cases of twinning and 

 numerous other biological factors, to say noth- 

 ing of faulty registration statistics, it is hard 

 to heUeve that the simple numerical relation 

 of triplets to twins can be more than the result 

 of the combination of numerous and as yet 

 unanalyzed forces. A knowledge of the fact 

 may, however, aid in the analysis. 



Chaeles Zeleny 



University of Illinois 



THE AMERICAN CHEMICAL SOCIETY 



SECTION OP SUGAK CHEMISTRY 



C. A. Browne, chairman 



I'red. J. Bates, secretary 



(Concluded) 



The sugar industry of Peru: Chas. A. Gamble. 



Electric oven for rapid moisture tests: G. L. 

 Spencer. This oven (patented August 3, 1920) is 

 a convenient arrangement for passing a rapid cur- 

 rent of heated air through a sample. The air is 

 drawn over a heating element, composed of a 

 spiraled nichrome wire coiled around a suitable 

 core, and thence through tiie sample contained in 

 a capsule, fitted with a gauze or metal filter-cloth 

 bottom. The temperature of the air is controlled 

 by a rheostat. Any substance through which a 

 current of hot air may be passed, without melting, 

 may be dried in this oven. Raw sugar may be 

 approximately dried in 3 minutes and to constant 

 weight in 10 minutes; 100 gram samples of cane 

 bagasse are dried in less than 60 minutes; cotton 

 saturated with water is dried in 10 minutes. 



