November i, 1894] 



NA TURE 



II 



and bark round the stump of a branch sawn off, is a reaction of 

 the organism on the action of an external influence. It seems, 

 therefore, fair to suppose that when our autlnr spealcs of a muti- 

 lation as an acquired character, he means the growth of the 

 organism consequent upon the mutilation. 15ut if so, a 

 ■difticulty appears to arise ; for the tendency to repair a wound is 

 heritable, and therefore it seems difficult to suppose that the 

 Professor would treat of it as not heritable. It may be said 

 that the parent has the actuality of the repair, the child 

 only the possibility of repairing ; bat this is, I suppose, all 

 that can ever be e.\pected of inheritance as applied to 

 -contingent reactions — i.e. reactions which only arise under 

 peculiar circumstances. There can, I suppose, be no more 

 ■characteristic heritable reaction than that of the pollen on the 

 ovule, and of the ovule on the pollen ; these reactions have taken 

 place in the parent plant, but in the offspring they are 

 originally potentialities, tendencies, contingencies, and they 

 are converted into facts in the event, and in the event only, 

 of the meeting of the pollen and the ovule. 



Or take, again, the secretion of the gastric juice in response to 

 the presence of food in the stomach. The parent has taken 

 food, and the reaction has taken place; but the infant inherits, 

 I suppose, the capacity to secrete and not the counterpart of the 

 actual secretion of the parent. 



It would seem that there is no essential difference between 

 these three cases. The parent transmits the power to repair 

 ji wound, but not the actual reparation of a wound : it t.ansmits 

 the power of fertilisation, but not the fertilised ovules : it 

 transmits the power to digest, but not the already secreted 

 gastric juice. 



The emphasis which Weismann has laid on the case of wounds 

 and mutilations would suggest that his doctrine might be thus 

 paraphrased : when an organism is endowed with a capacity, or, 

 to use his word, a predisposition, to react in response to given 

 stimuli, and has so reacted — then what the organism transmits 

 k> its progeny is the capacity or predisposition, and not the 

 actual result of the reaction. 



It is impossible to doubt that some characters of an organism 

 are hereditary ; that others are not, and that the ascertainment 

 ■of the dividing line between the two classes is of the highest 

 moment to the study of biology ; and to Weismann we owe a 

 ■debt of gratitude for having called pointed attention to this 

 matter. I have at times been tempted to wish that men of 

 science had applied themselves to ascertain the two categories 

 of characters, and then by a careful induction had proceeded 

 •to learn the law of heredity without regard to any hypothesis 

 or theory — without reference to germ or soma. But this is 

 not the course which in fact has been taken ; and therefore it 

 «eems highly necessary to inquire what is the precise meaning 

 of the terms of the proposition affirmed by the one side, and 

 denied by the other. 



I conclude this, I fear, too lengthy paper with two questions : 

 i.\) Are the conditions which I have suggested as essential to a 

 good definition correct ones ; if not, in what are they erroneous? 

 (2) What is the true definition of the words "acquired 

 characters '' in the present controversy which satisfies these 

 <:onditions? Edw. Fry. 



Discontinuous Motion. 



The old theory of the motion of solid bodies through a 

 ftictionless liquid supposed that the li(iuid flowed according 

 to the electrical law of flow. This theory was found to be un- 

 satisfactory, because it makes the pressure negative when the 

 velocity of the solid exceeds a certain critical value. 



The theory of discontinuous motion removes the above 

 objection, but is open to others of a different kind. Assuming 

 for the sake of argument that the two theories give correct 

 results when the velocity of the moving solid is respectively 

 Jess and greater than the critical value, the theory of discon- 

 tinuous motion ought to be capable of explaining the transition 

 from one kind of motion to the other, and how and why it is 

 possible for a vortex sheet to be called into existence when the 

 critical value of the velocity is exceeded. 



Although vortex sheets and other motions involving molecular 

 rotation cannot be generated in a frictionless liquid by a con- 

 servative system of forces or by operations perlormed on the 

 ■boundary, yet it is easy enough to produce such motions by 

 ordinary mechanical agencies. If a mixture of ice and water 

 be stirred up and the ice allowed to melt, the liquid will 



MO. 1305, VOL. 51] 



acquire molecular rotation owing to the presence of the particles 

 of melted ice, even though it is absolutely devoid of viscosity. 

 So also if liquid at rest were separated by an indefinitely thin 

 horizontal plate, and the upper liquid were set in motion with 

 horizontal velocity V and the plate were removed, the surface 

 of separation would be a vortex sheet. But the production of 

 motions of this kind requires methods of a somewhat artific iai 

 character, and it is difficult to see how they could be set up by 

 a solid whose velocity is allowed to increase gradually from zero 

 to some magnitude greater than the critical value. In fact, I 

 entertain very little doubt that the final motion of the liquid 

 would be quite different from what the theory of discontinuous 

 motion would indicate. 



There is, however, a further point, for there are strong 

 grounds for believing that vortex sheets are unstable. No 

 general proof of this proposition appears as yet to have been 

 given, but in every case that has been examined the theorem 

 has been found to be true, (i.) when the liquids on either side of 

 the sheet are identical, (ii.) when the densities of the two 

 liquids are different, but no bodily forces such as gravity and 

 the like are in action. If, therefore, steady discontinuous 

 motion existed at any particular instant, the probabilities are 

 that the motion would be unstable, and the region of dead 

 water in the rear of the moving solid would break up and be 

 changed into a region of turbulent motion. The pressure in 

 the rear of the solid due to this turbulent motion would be 

 different from that of the dead water, and it is therefore not 

 surprising that the theory of discontinuous motion should 

 furnish results which do not agree very well with experiment. 



We must also recollect that a frictionless liquid is an ideal 

 substance which does not exist in nature. .\ll fluids are more 

 or less viscous ; and it is just at the point where the pressure 

 would otherwise tend to vanish and change sign that we should 

 anticipate the effect of viscosity would appear, and prevent 

 this stale of things from taking place ; and I believe that many 

 of the difficulties which have arisen in connection with this sub- 

 ject are due to the fact that the effect of viscosity has been over- 

 looked. A vortex sheet cannot exist in a viscous liquid ; and if by 

 any artificial means one were produced, it would immediately dis- 

 appear, and molecular rotation would be propagated into the sur- 

 rounding liquid. On the other hand, in a viscous liquid, molecu- 

 lar rotation requires no artificial means for its production ; for a 

 viscous liquid cannot move without molecular rotation, except 

 in the single case in which the liquid moves like a rigid body 

 having a motion of translation alone. In all other cases, il 

 irrotational motion existed at any particular in -tant. the motion 

 would immediately cease to be so, and molecular rotation 

 would instantaneously be generated. 



Unfortunately the equations of motion of a viscous liquid 

 are so intractable that very little progress has been made in 

 applying them to the solution of hydrodynamical problems. 

 By means of Oberbeck's solution [Borchariit'i J otirnal, yo\. 

 Ixxx. ) for the steady motion of translation of an ellipsoid in 

 a viscous liquid, it can be shown that the above difficulties do 

 not arise when viscosity is taken into account ; but since the 

 integration of the equations of motion proceeds upon the 

 assumption that the squares and products of the velocities may 

 be neglected, the solution is inapplicable except in the case of 

 slow motions like those produced by the small oscillations of 

 pendulums. The solution gives no information as to what will 

 happen when a disc is moving through a liquid with a velocity 

 of several feet per second. A. B. Basset. 



Fledborough Hall, Ilolyport, Berks. 



Capacity for Heat. 



In the course of some writing upon which I am now engaged, 

 I have constantly to refer to the capacity for heat of certain 

 substances as compared ■.villi the capacity for heat of an cqtuil 

 volume of water. 



The phrase given in italics is a most clumsy one, but I know 

 not how (accurately) to convey the same idea in a shorter way. 

 " Cafacilyfor heat of unit volume" has been suggested ; but I 

 think that a little reflection will show that it does not express 

 accurately the exact meaning. 



Specific Heat X Specific Craiity gives the numerical value 



required, but cannot be regarded as a definition. There can be 



no doubt but that a concise expression is wanted. In calori- 



metry it is often of greater importance to the experimenter to 



onsider the capacity for heat of volumes rather than the capacity 



