WCP4137

[1]^1,2

Experiments for testing heredity of acquired character With Plants.

1. Effect of wind.

Two identical sets of ^perennial plants. The set grown for successive generations fully exposed to wind, both natural and artificial — the other set grown in as still an atmosphere as possible consistent with health.

When a decided effect of the different conditions is perceptible, let seeds of both sets be grown under identical conditions, and it will be seen whether the individual difference is transmitted.

2. The effects of drought and moisture, could be tested in the same way.

3. The effects of diversity of soil — chalky, sandy, clayey,- peaty,- might be similarly tried.

4. The effects of light and shade, might be easily tried.

[2]³Remarks.

Experiments made with a variety of species as here suggested would test the theory in two ways;- firstly, by showing whether any increased effect was produced in successive generations, and, secondly, by ascertaining whether the effects produced, leading to a diversity in the two sets, continued to any degree when the seeds of both were grown under identical conditions.

Of course, selection of any kind must be rigourously excluded. The seed for each successive sowing must be taken by some mechanical method that would give a fair sample of the total amount of seed produced. Plants must be chosen which are vigorous enough to grow under the varied conditions without injury to their health or powers of reproduction; and considerable numbers must be grown so as to eliminate the effects of accident and to render a fair [3]⁴sampling of the whole crop possible.

If any decided effect in the individual plants is produced by the diverse conditions of growth here suggested, ^{acting on individuals for} several ^years, there would not appear to be any great difficulty in deciding whether those effects were or were not transmitted to succeeding generations.

[4]⁵

Experiments for testing heredity of acquired Characters

With Animals.

1. Two sets of common dovecot pigeons might be exposed to different conditions, one set being allowed freely to fly about, the other set being confined in a spacious yard netted over at two or three feet from the ground so as to prevent flight. Birds reared and kept under these different conditions for several years would probably exhibit differences in the development of the bones and muscles of the wings and legs respectively. After these differences become distinctly perceptible and capable of measurement, the offspring of the two sets should be reared under identical conditions when it would be seen whether the effects produced in the individuals were transmitted to their offspring.

2. Ducks might be reared and kept for several years without access to water except for drinking & washing, another identical set having ample swimming space. Some ^acquired individual differences might arise between the [5]⁶two sets as regards the webbed feet or the development of the muscles used in swimming. If ^{any such changes could be} detected heredity could be tested as in the case of the pigeons.

3. Poultry, pigeons, ducks, pigs, rabbits &c. could be fed on two kinds of food, the one very nutritious the other comparatively innutritious, but so arranged as both to be compatible with health. After several years of the treatment, beginning with very young animals, a considerable difference would probably be found in the average size & weight of the two sets. When this was clearly established the offspring of the two sets, if exposed to identical treatment, would show whether the effects produced on their parents were hereditary.

[6]⁷4. ex. Similar experiments to the last could be tried with Lepidoptera⁸ of various species. The particular kinds or quantity of food ^{of the larvae} being regulated so as to produce broods larger or smaller than the average.

5. In other cases the larvae might be exposed to low temperatures so as to produce melanism. and both these By breeding from considerable numbers of insects thus treated and subjecting these ^{and a corresponding set bred normally} to identical conditions it would be seen whether the modification produced in the individuals was transmitted to the offspring.

6. (see over)

In these experiments with animals, as in those with plants, considerable numbers of individuals must be employed, and at every stage of the process selection must be rigidly excluded.

[7]⁹6. A more expensive but a good and practicable experiment would be with, say ^{three or} half-a-dozen half wild horses such as the Mexican mustangs (or better still some really wild horses.) These should be bred till a lot of, say, twenty were produced, these divided into two lots of equal quality should thenceforth be systematically trained, the one lot as ^{hunters on} racers the other lot as carthorses these being daily used in drawing weight at walking pace and never allowed to run. After five or six years these should be bred from to see if the offspring exhibited any differences in the direction of cart-horse or racer respectively.

[8]¹⁰Experiments on Instinct

1 An important series of Experiments might be carried on as to the alleged instincts of the higher animals.

1. The supposed nest-building instinct of birds could be tested by means of a wired enclosure of considerable extent with bushes and shrubbery, into which young birds could be turned loose, after having been reared in such a way that they could not have seen the proper nest of their species. Half a dozen experiments of this kind, with the nest built by the birds wholly from their "inner consciousness" placed by the side of the typical nest of the species, would do more towards the settlement of this vexed question than all that has yet been written on it.

2. The alleged "direction instinct" of dogs, cats, and other animals could be tested by a few experiments which have never yet been made. Dogs or cats whose knowledge of the locality has been restricted to a mile or less from their place of birth, should [9]¹¹be taken thoroughly blindfolded, and by a very circuitous route, to a spot from 3 to 10 miles distant according to the nature of the country, and their proceedings carefully watched and recorded by some person not known to them till they either found their way home or gave up the attempt to go home. We should thus learn their mode of action, what senses or faculties were used, and what proportions of the species were able to find their way home across a completely unknown intervening country. We should learn whether any special instinct was involved, or whether, (as I think more probably) all the recorded cases are the result of the use of the ordinary senses faculties of sight, smell, and memory, combined with a certain amount of good luck.

3. Another important set of experiments would be the repetition and extension of those made by the late Mr. Spalding¹² on newly— [10]¹³— born animals. It might thus be ascertained whether the knowledge or experience of the parent is inherited, or merely those nervous and muscular co-ordinations which render some sensations pleasurable other disagreeable or alarming, and which produce corresponding reflex actions which simulate the result of experience.

[11]¹⁴Experiments on the fertility of Hybrids.

Animals.

In order to arrive at any definite and trustworthy results it will necessary ^(a)- to choose the species of such genera as are known to breed freely in confinement. (b) Several closely allied species should be chosen and all the possible crosses made between them. (c) The hybrid offspring produced ^{by each cross} should be segregated, and so treated as to give them every opportunity of breeding. (d) A considerable number of individuals — 6 to 10 ^{at least}— of each sex should be obtained before experimenting with any of the species.

Among the animals that might probably serve for these experiments are the following: —

Mammalia. — Species of Dasyprocta (Agoutis) and Cavia (Guinea pigs); Lepus (Hare and Rabbit); Gazella, (Gazelles); Capra (Goats, Iberes &c.) Ovis (wild Sheep); Cervus (Deer). In the three last groups there are many allied but distinct species which offer excellent materials for Experiment in hybridisation.

[12]¹⁵Birds. — The families Anatidae (Ducks and Geese) and Phasianidea (Partridges, Pheasants, Quails and Jungle fowl) offer perhaps the best material for experiment among birds, as they breed rapidly, and many breed well in confinement, and some are known to hybridise.

Insects. — Experiments might perhaps succeed in the hybridisation of some of the allied species of Bombycidae ^{a family} which includes the silk-worm moths.

Plants.

A very wide field is open for experiment as to the fertility of hybrids among plants. The numerous experiments hitherto made have not had this special object in view. In some cases the experimenters have held that the fertility of hybrids proved the non-specific distinctness [13]¹⁶of the parents. Others have been more interested in showing that very unlike plants, often even when belonging to distinct genera, could be hybridised. Others again have sought only to improve the race for horticultural purposes.

A systematic series of experiments in crossing the various species of genera which are usually easy of cultivation, and such as have usually numerous well-marked though closely allied species, would lead to most interesting results. The results of crossing representative species from different continents or from opposite sides of a continent, and of those which inhabit a continent and a continental or oceanic island, would be of the greatest value, as indicating whether some difficulty in producing hybrids and considerable or complete infertility of the hybrid offspring is a constant character of species, or is in any way dependent on their having been differentiated in the same or in [14]¹⁷isolated areas.

[1]¹ [p. 130]

Parkstone, Dorset, Feby. 7th, 1891.

My Dear Mr Galton, On receipt of your interesting letter I sat down & jotted the enclosed notes of the kind of experiments that it seems to me would test the theory of heredity or non-heredity of individually-acquired characters. Also a few as to fertility or sterility of hybrids, & as to the real nature of some of the supposed instincts of the higher animals. I do not myself see much difficulty in carrying out any of these, but then I am not an experimenter as you are. Still, I shall be glad to know exactly where the difficulty or insufficiency lies. If these, or any modifications of them, would be valuable & to the point, it seems to me that the mere keeping the plants and animals in health & properly isolated would fully occupy the keeper or keepers of the farm,—while the actual experiments —the deciding on the separation without selection of the various lots to experiment with,—which should be crossed & when, and other such matters, would recur only at considerable intervals & could be supervised by the members of the Committee, or some of them, by means of, say, a weekly inspection.

I have limited my notes to three points in which I feel most interest, but of course experiments in variation such as Mr Merrifield is carrying on for you, could be added to any extent if there were any danger of the keepers having too little to do!

All the experiments I suggest would require considerable numbers of individuals to be kept healthy and to be largely increased by breeding,—and they would all have to be continued during several years depending on the duration of life of the various species experimented with.

My wife and I are in pretty good health & beg to be kindly remembered to Mrs Galton. As everybody seems to come to Bournemouth we shall hope some day to have a call from you.

Yours very faithfully, Alfred R. Wallace.

F. Galton, Esq., F.R.S.