TRESEXT rROHLKMS IN EVOLUTION AND HEREDITY. 1529 



niosl it'iiiotc j)('ii()<l. llic (lisuox cry ol';i similar iiiusclr may be iimmcIv a 

 coincidence; it is by no means a jtroot' ot" reversion. 



The lirst test of reversion is llierelbre the anatomy ol the ata\us 

 and this is dei'ived i)artly from the i>aLeontolo<;ical record oC the 

 ])riinates, partly from the law of diveri^ence, viz., that featnres which 

 are common to all the livinii" ])riinates Avere ])robal)ly also found in the 

 stem form which uave rise to man; linally, from the comparative anat- 

 omy of the li\ iuji' antliroi»oidea. 



The second test is whether a strncture passes the limits of reversion 

 as determined by cases of ata\'ism in which theiecan be no reasonable 

 doubt. Two of these phenomena have recently been discussed, which 

 seem to extend the ])ossibi]ities of reversion back to structmes which 

 were lost at a very remote period. I refer to ]>apers by Williams and 

 Howies. Williams* has analyzed lf>(> rec(trded <'ases of polymastism; 

 he finds that sui)ernumerary nii)])les of souk^ form occur in two per 

 cent, and that in all except four of the cases examined the anomalies, 

 tested by position, etc., su]iport the reversion hypothesis. In the 

 living" lemurs, whiidi form a ]»ersistent primitive group of monkeys, we 

 find that the transition from i)olymastism to bimastism is now in 

 progress by the degeneration of the abdominal and inguinal nii)ides; it 

 is fair to assume that the higher monkeys also lost their abdominal 

 nipples at a ]>rimitive stage of development, and therefore that cases 

 of multiple nipples indicate reversion to a Lower Eocene condition! 

 Howes t has recently comi^leted a most interesting study of the ''iu- 

 tranarial ei)iglottis," or cases in which the epiglottis is carried up into 

 the posterior mires, as in young marsupials and some cetacea, to sub- 

 serve direct narial respiration. This has now been ol)served to occur 

 by reversion in all orders of mammals, including the monkeys and 

 lemurs. One case has also been reported by Sutton of its occurrence 

 in a human fetus. This is api)arently a human revei'sion to a structure 

 much older than the age of the lemurs. 



The third test is the inverse latio to time. It woidd seem, <i priori^ 

 that the i>ercentageofrecunencei)f atavistic structures should decrease 

 as the extent of time elai)sing since the structure disai)])eare(l increases. 

 This law is api»arently established in the case of the condylar and 

 intercondylar foramina, and if we examine all the ])ercentages which 

 have been established we see at once that they bear a ratio to time; 

 <;om])are the relative fie(|uenc3^ of the isc,hio-i)ul)ic (.">() per cent), dorso- 

 epitrochlearis (5 per cent), and levator-davicuhr (!••><' p<'i' ecnt) mus- 

 cles with the periods which have elapsed sinci^ their past service. This 

 is why it is so imj)ortant to establish ]»erc('ntages for all our atavistic 

 organs; fidler statistics wall not only bear upon heicdity, but I can 

 conceive of their application to the extremely diflicnlt jjroblem of 

 estinniting geological time. \N'e must, of course, establish as a stand- 



^ Jonrnnl (i/ Aiuiliniiji luid riinnioliH/i/, 18!ll. p. 224. 

 t Ibid, l!S8!t, ]). 5X7. 



