Arber — Atavism and Law Irreversibility. 27 



Aet. IV. — On Atavism and the Law of Irreversibility? 

 by Agnes Arber, D.Sc. 



In a paper recently read before the Linnean Society, 1 

 I have drawn attention to a certain minor principle of 

 evolution — the 'Law of Loss' — deduced from considera- 

 tions relating to the comparative morphology of living 

 plants. I have defined the expression 'Law of Loss' as 

 indicating the "general rule that a structure or organ 

 once lost in the course of phylogeny can never be 

 regained; if the organism subsequently has occasion to 

 replace it, it cannot be reproduced, but must be con- 

 structed afresh in some different mode." 2 The 'Law of 

 Loss' corresponds in fact to a part of the broader prin- 

 ciple recognized by Dollo 3 under the name of the 'Law of 

 Irreversibility, ' though I was unacquainted with the 

 Belgian Professor's work on the subject at the time when 

 I formulated the Law of Loss. Since the paper in ques- 

 tion was written, my attention has been called to a pas- 

 sage in Professor J. Bretland Farmer's "Plant Life," 

 in which a principle is enunciated also covering part of 

 the ground of Dollo 's Law. In the chapter dealing with 

 the relating of plants to water, the following sentences 

 occur: "It is a singular fact that when a species or race 

 has once exhibited a tendency towards the loss or atro- 

 phy of an organ, e. g. the leaf, the descendants commonly 

 appear unable to check it. If any of them vary in such 

 a way as to increase their green surface, this is effected 

 not by enlarging their diminished leaves, but by flatten- 

 ing and specialising some other organ." 4 We may — 

 perhaps rather crudely — interpret this suggestion as 

 indicating that the path of degeneration represents a 

 downward slope on which progress, once begun, con- 



1 The 'Law of Loss' in Evolution. Eead 7th Nov. 1918 (to appear in 

 Proc. Linn. Soc. Lond., 1919). 



2 This law applies only to actual losses and not to apparent losses due 

 to the interpolation of inhibiting factors; the necessity for this distinction 

 has been pointed out to me by Miss E. K. Saunders. 



3 Dollo, L., Les lois de 1 'evolution. Bull. Soc. Beige de geol. paleont. 

 hydrol., 7, Proces-verb., Seance du 25 juillet, 1893, pp. 164-166. See also 

 a recent paper by Dr. B. Petronievics in which Dollo 's memoirs on the sub- 

 ject are analysed in detail (Sur la loi de 1 'evolution irreversible, Science 

 Progress, vol. 13, pp. 406-419, Jan. 1919). 



4 Another passage bearing on the same subject may also be cited (p. 113) : 

 "And there are a very large number of instances, of the most varied kind, 

 which indicate that when an organism has once modified its constitution 

 so as to exhibit any special trend, the chances are all in favour of advance 

 along the new lines, and very slightly indeed in favour of a return to the old 

 ones. ' ' 



