to the miller the quality of wheat depends on three chief factors, the percentage of dirt, weed seeds, and other impurities, the percentage of water in the sample, and a complex and somewhat ill-defined character commonly called strength. 52

with the methods of growing, cleaning and thrashing wheat practised in great britain, practically clean samples are produced, and home grown wheat is therefore on the whole fairly free from impurities. this is, however, far from the case with foreign wheats, many of which arrive at the english ports in an extremely dirty condition. they are purchased by millers subject to a deduction from the price for impurities above the standard percentage which is allowed. the purchase is usually made before the cargo is unloaded. official samples are taken during the unloading in which the percentage of impurities is determined, and the deduction, if any, estimated.

the percentage of water, the natural moisture as it is usually called, varies greatly in the wheats of different countries. in home grown wheats it is usually 16 per cent., but in very dry seasons it may be much lower, and in wet seasons it may rise to 18 per cent. foreign wheats are usually considerably drier than home grown wheats. in russian wheats 12 per cent. is about the average, and that too is about the figure for many of the wheats from canada, the states, argentina, and parts of australia. indian wheats sometimes contain less than 10 per cent. this is also about the percentage in the wheats of the arid lands on the pacific coast and in australia. these figures show that home grown wheats often contain as much as 5 per cent. more water than the foreign 53 wheats imported from the more arid countries. the more water a wheat contains the less flour it will yield in the mill. consequently the less its value to the miller. a difference of 5 per cent. of natural moisture means a difference in price of from 1s. 6d. to 2s. per quarter in favour of the drier foreign wheats. this is one of the reasons why foreign wheats command a higher price than those grown in this country.

turning to the third factor which determines the quality of wheat from the miller’s point of view, we may for the present define strength as the capacity for making bread which suits the public taste of the present day. we shall discuss this point more fully when we deal with the baking of bread. at present the only generally accepted method of determining the strength of a sample of wheat is to mill it and bake it, usually into cottage loaves. the strength of the wheat is then determined from their size, shape, texture, and general appearance. a really strong flour makes a large, well risen loaf of uniformly porous texture. wheats lacking in strength are known as weak. a weak wheat makes a small flat loaf. in order to give a numerical expression to the varying degrees of strength met with in different wheats, the home grown wheat committee of the national association of british and irish millers have adopted a scale as the result of many thousand milling 54 and baking tests. on their scale the strength of the best wheat imported from canada, graded as no. 1 manitoban, or from the states graded as no. 1 hard spring, is taken as 100, that of the well-known grade of flour known as london households as 80, and that of the ordinary varieties of home grown 55 wheat, such as square head’s master, browick, stand up, etc., as 65. the strength of most foreign wheats falls within these limits. thus the strength of ghirka wheat from russia is about 85, of choice white karachi from india 75, of plate river wheat from the argentine 80, etc. the strongest of all wheats is grown in certain districts in hungary. it is marked above 100 on the scale, but it is not used for bread making. the soft wheats from the more arid regions in australia and the states are usually weaker than average home grown samples, and are marked at 60. rivet or cone wheat, a heavy cropping bearded variety much grown by small holders,—since the sparrow, which would ruin small plots of any other variety, seems to dislike rivet, possibly on account of its beard,—is the weakest of all wheats, and is only marked at 20, which means that bread baked 56 from rivet flour alone would be practically unsaleable. rivet wheat finds a ready sale, however, for making certain kinds of biscuits.

in order to make flour which will bake bread to suit the taste of the general public of the present day, the miller finds it necessary to include in the mixture or blend of wheats which he grinds a certain proportion of strong wheats such as canadian, american, or russian. the quantity of strong wheat available is limited. consequently strong wheat commands a relatively high price. the average difference in price of say no. 1 manitoban and home grown wheat is about 5s. per quarter. it is possible of course that the public taste in bread may change, and damp close textured bread may become fashionable. in this case no doubt the difference in price would disappear. under present conditions the necessity of including in his grinding mixture a considerable proportion of strong foreign wheat is a distinct handicap against the inland miller as compared with the port miller. the latter gets his foreign wheat direct from the ship in which it is imported, whilst the former has to pay railway carriage from the port to his mill. the question naturally arises—is it not possible to grow strong wheats at home and sell them to the inland miller?

this question has been definitely answered by the work of the home grown wheat committee 57 during the last 12 years. the committee collected strong wheats from every country where they are produced, and grew them in england. from the first crop they picked out single plants of every type represented in the mixed produce, for strong wheats as imported are usually grades and not pure varieties. from the single plants they have established pure strains of which they have grown enough to mill and bake. from most of the strong wheats they were unable to find any strain which would produce strong wheat in england. thus the strong wheat of hungary when grown in england was no stronger than any of the ordinary typical home grown wheats. but from the strong wheat of canada was isolated the variety known as red fife, which makes up a very large proportion of the higher grades of american and canadian wheats, and this variety when grown in england was found to continue to produce wheat as strong as the best canadian. year after year it has been grown here, and when milled and baked its strength has been found to be 100 or thereabouts on the scale above described. finally it was found that a strain of red fife which had been brought over from canada 20 years ago, and grown continuously in the western counties ever since, under the name of cook’s wonder, was still producing wheat which when ground and baked possessed a strength of about 100. thus it was conclusively proved that 58 in the case of red fife at any rate the english climate was capable of producing really strong wheat. the strength of hungarian and russian wheats appear to be dependent on the climate of those countries. red fife, however, produces strong wheat wherever it is grown. it is interesting to note that this variety although first exploited in canada and the states is really of european origin. it was taken out to canada by an enterprising scotchman called fife in a mixed sample of dantzig wheat. he grew it for some time and distributed the seed. pure strains have from time to time been selected by the american and canadian experiment stations.

but the discovery that red fife would produce strong wheat in england by no means solved the problem, for when the home grown wheat committee distributed seed of their pure strain of that variety for extended testing throughout the country, it was soon found to be only a poor yielder except in a few districts. a yield of three quarters of strong grain, even if it makes 40s. per quarter on the market, only gives to the farmer a return of ￡6 per acre, as compared with a return of nearly ￡8 from 4? quarters of weak grain worth 35s. per quarter, which can usually be obtained by growing square head’s master, or some other standard variety.

it was at this point that mendel’s discoveries came to the rescue. working on the mendelian lines 59 already explained, biffen at cambridge crossed red fife with many of the best english varieties. from one of the crosses he was able to isolate a new variety in which are combined the strength of red fife and the vigour and cropping power of the english parent. this variety, known as burgoyne’s fife, has been grown and distributed by members of the millers’ association. in 1911 on the average of 28 separate trials it yielded 38 bushels per acre, which is well above the average of the best english varieties. it has been repeatedly milled and baked, and its strength is between 90 and 100, practically the same as that of red fife. it has been awarded many prizes at agricultural shows for quality, and it commands on markets where the local millers have found out its baking qualities about the same price as the best 60 foreign strong wheats, that is to say from 4s. to 5s. per quarter more than the average price of home grown wheat. taking a fair average yield of wheat as four quarters per acre, burgoyne’s fife gives to the farmer an increased return over the ordinary varieties of about 16s. per acre. the introduction of such a variety makes the production of strong wheat in england a practicable reality, and will be a boon both to the farmer and to the inland miller. it is likely too that the possibility of obtaining a better return per acre will induce farmers to grow more wheat. anything that tends to increase the production of home grown wheat and makes great britain less dependent on foreign supplies is a national asset of the greatest value.

it is of the greatest importance to the miller that he should be able to determine the strength of the wheats he buys. obviously the method mentioned above, which entails milling enough of the sample to enable him to bake a batch of bread, is far too lengthy to be of use in assessing the value of a sample with a view to purchase. the common practice is for the miller or corn merchant to buy on the reputation of the various grades of wheat, which he confirms by inspection of the sample. strength is usually associated with certain external characters which can readily be judged by the eye of the practised wheat buyer. strong wheats are usually red in colour, their 61 skin is thin and brittle, the grain is usually rather small, and has a very characteristic horny almost translucent appearance. the grains are extremely hard and brittle, and when broken the inside looks flinty. on chewing a few grains the starch is removed and there remains in the mouth a small pellet of gluten, which is tough and elastic like rubber, but not sticky.

weak wheats as a rule possess none of these characters. their colour may be either red or white, their skin is commonly thick and tough, the grain is usually large and plump, and often has an opaque mealy appearance. it is soft and breaks easily, and the inside is white, soft and mealy. very little gluten can be separated from it by chewing, and that little is much less tough and elastic than the gluten of a strong wheat.

these characters, however, are on the whole less reliable than the reputation of the grade of wheat under consideration. to make a reliable estimate of strength from inspection of a sample of wheat requires a natural gift cultivated by continual practice. even the best commercial judges of wheat have been known to be deceived by a sample of white wheat which subsequent milling and baking tests showed to possess the highest strength. the mistake was no doubt due to the great rarity of strength among white wheats. this rarity will doubtless soon 62 disappear now that a pure strain of white fife has been isolated and shown to possess strength quite equal to that of red fife. sometimes too the ordinary home grown varieties produce most deceptive samples which show all the external characters of strong wheats. such samples, however, on milling and baking are invariably found to possess the usual strength of home grown wheat, about 65 on the scale. these considerations show the great need of a scientific method of measuring strength, which can be carried out rapidly and on a small sample of grain. this need is felt at the present time not only by the miller and the merchant, but by the wheat breeder. for instance, in picking out the plants possessing strong grain from cultures of the second generation after making his crosses, the plant breeder up to the present has had to rely on inspection by eye, and on the separation of gluten by chewing, for a single plant obviously cannot yield enough grain to mill and bake. this fact no doubt explains the differences of opinion among plant breeders on the inheritance of strength, for it is not every one who can acquire the power of judging wheat accurately by his senses. such a faculty is a personal gift, and is at best apt to fail at times.

the search for a rapid and accurate method of measuring strength has for many years attracted the attention of investigators. as might be expected 63 most of the investigations have centred round the gluten, for as mentioned above the gluten of a strong wheat is much more tough and elastic than that of a weak wheat. gluten is a characteristic constituent of all wheats, and it is the presence of gluten which gives to wheat flour the power of making bread. the other cereals, barley, oats, maize and rice are very similar to wheat in their general chemical composition, but they do not contain gluten. consequently they cannot make bread.

in making bread flour is mixed with water and yeast. the yeast feeds on the small quantity of sugar contained in the flour, fermenting it and forming from it alcohol and carbon dioxide gas. the gluten being coherent and tough is blown into numberless small bubbles by the gas, which is thus retained inside the bread. on baking, the high temperature of the oven fixes these bubbles by drying and hardening their walls, and the bread is thus endowed with its characteristic porous structure. if a cereal meal devoid of gluten is mixed with water and yeast, fermentation will take place with formation of gas, but the gas will escape at once, and the product will be solid and not porous. evidently from the baking point of view gluten is of the greatest importance. one of the most obvious methods that have been suggested for estimating the strength of wheat depends on the estimation of the 64 percentage of gluten contained in the flour. the method has not turned out very successfully, for strength seems to depend rather on the quality than on the quantity of gluten in the wheat. much attention has been given to the study of the causes of the varying quality of the gluten of different wheats. gluten for instance has been shown to be a mixture of two substances, gliadin and glutenin, and the suggestion has been made that its varying properties are dependent on the varying proportions of these two substances present in different samples. this suggestion however failed to solve the problem.

after seven years of investigation the author has worked out the following theory of the strength of wheat flours, which has finally enabled him to devise a method which promises to be both accurate and rapid, and to require so little flour that it can readily be used by the wheat breeder to determine the strength of the grain in a single ear. it has already been mentioned that a strong wheat is one that will make a large loaf of good shape and texture. the strength of a wheat may therefore be defined as the power of making a large loaf of good shape and texture. evidently strength is a complex of at least two factors, size and shape, which are likely to be quite independent of each other. not infrequently, for instance, wheats are met with which make large loaves of bad shape, or on the other hand, small 65 loaves of good shape. probably therefore the size of the loaf depends on one factor, the shape on another; and the failure of the many attempts to devise a method of estimating strength have been caused by the impossibility of measuring the product of two independent factors by one measurement.

it seemed a feasible idea that the size of the loaf might depend on the volume of gas formed when yeast was mixed with different flours. on mixing different flours with water and yeast it was found that for the first two or three hours they all gave off gas at about the same rate. the reason of this is that all flours contain about the same amount of sugar, approximately one per cent., so that at the beginning of the bread fermentation all flours provide the yeast with about the same amount of sugar for food. but this small amount of sugar is soon exhausted, and for its subsequent growth the yeast is dependent on the transformation of some of the starch of the flour into sugar. wheat like many other seeds contains a ferment or enzyme called diastase, which has the power of changing starch into sugar, and the activity of this ferment varies greatly in different wheats. the more active the ferment in a flour the more rapid the formation of sugar. consequently the more rapidly the yeast will grow, and the greater will be the volume of gas produced in the later stages of fermentation in the 66 dough. as a rule it is not practicable to get the dough moulded into loaves and put into the oven before it has been fermenting for about six or eight hours. if the flour possesses an active ferment it will still be rapidly forming gas at the end of this time, and the loaf will go into the oven distended with gas under pressure from the elasticity of the gluten which forms the walls of the bubbles. the heat of the oven will cause each gas bubble to expand, and a large loaf will be the result. if the ferment of the flour is of low activity it will not be able to keep the yeast supplied with all the sugar it needs, the volume of gas formed in the later stages of the fermentation of the dough will be small, the dough will go into the oven without any pressure of gas inside it, little expansion will take place as the temperature rises, and a small loaf will be produced.

from these facts it is quite easy to devise a method of estimating how large a loaf any given flour will produce. the following method is that used by the author. a small quantity of the flour, usually 20 grams, is weighed out and put into a wide mouthed bottle. a flask of water is warmed to 40° c., of this 100 c.c. is measured out, and into it 2? grams of compressed yeast is intimately mixed, 20 c.c. of the mixture being added to the 20 grams of flour in the bottle. the flour and yeast-water are then mixed 67 into a cream by stirring with a glass rod. the bottle is then placed in a vessel of water which is kept by a small flame at 35° c. the bottle is connected to an apparatus for measuring gas, and the volume of gas given off every hour is recorded. as already mentioned all flours give off about the same volume of gas during the first three hours. after this length of time the volume of gas given off per hour varies greatly with different flours. thus a flour which will bake a large loaf gives off under the conditions above described about 20 c.c. of gas during the sixth hour of fermentation, whilst a flour which bakes a small tight loaf gives off during the sixth hour of fermentation only about 5 c.c. of gas.

having devised a feasible method of estimating how large a loaf any given flour will make, the problem of the shape and texture still remains. previous investigators had exhausted almost every possible chemical property of gluten in their search for a method of estimating strength. the author therefore determined to study its physical properties. now gluten is what is known as a colloid substance, like albumen the chief constituent of white of egg, casein the substance which separates when milk is curdled, or clay which is a well known constituent of heavy soils. such colloid substances can scarcely be said to possess definite physical properties of their own, for their properties vary so largely with their 68 surroundings. the white of a fresh egg is a thick glairy liquid. on heating it becomes a white opaque solid, and the addition of certain acids produces a similar change in its properties. casein exists in fresh milk in solution. the addition of a few drops of acid causes it to separate as finely divided curd. if, however, the milk is warmed before the acid is added the casein separates as a sticky coherent mass. every farmer knows that lime improves the texture of soils containing much clay, because the lime causes the clay to lose its sticky cohesive nature.

such instances show that the properties of colloid substances are profoundly modified by the presence of chemical substances. wheat, like almost all plant substances, is slightly acid, and the degree of acidity varies in different samples. accordingly the effect of acids on the physical properties of gluten was investigated, and it was found that by placing bits of gluten in pure water and in acid of varying concentration it could be made to assume any consistency from a state of division so fine that the separate particles could not be seen, except by noticing that their presence made the water milky, to a tough coherent mass almost like indiarubber (fig. 11). it was found, however, that the concentration of acid in the wheat grain was never great enough to make the gluten really coherent.

but wheat contains also varying proportions of such salts as chlorides, sulphates and phosphates, which are soluble in water, and the action of such salts on gluten was next tried. it was at once found that these salts in the same concentration as they exist in the wheat grain were capable of making gluten coherent, but that the kind of coherence produced was peculiar to each salt. phosphates produce a tough and elastic gluten such as is found in the strongest wheats. chlorides and sulphates on the other hand make gluten hard and brittle, like the gluten of a very weak wheat (fig. 12).

the next step was to make chemical analyses to find out the amount of soluble salts in different wheats. strong wheats of the fife class were found to contain not less than 1 part of soluble phosphate in 1000 parts of wheat, whilst rivet wheat, the weakest wheat that comes on the market, contained only half that amount. rivet, however, was found to be comparatively rich in soluble chlorides and sulphates, which are present in very small amounts in strong wheats of the fife class. ordinary english wheats resemble rivet, but they contain rather more phosphate and rather less chlorides and sulphates. after making a great many analyses it was found that the amount of soluble phosphate in a wheat was a very good index of the shape and texture of the loaf which it would make. the toughness and elasticity 71 72 of the gluten no doubt depend on the concentration of the soluble phosphate in the wheat grain, the more the soluble phosphate the tougher and more elastic the gluten, and a tough and elastic gluten holds the loaf in shape as it expands in the oven, and prevents the small bubbles of gas running together into large holes and spoiling the texture.

these facts suggest at once a method for estimating the shape and texture of the loaf which can be made from any given sample of wheat. an analysis showing the amount of soluble phosphate in the sample should give the desired information. but unfortunately such an analysis is not an easy one to make, and requires a considerable quantity of flour. in making these analyses it was noticed that when the flours were shaken with water to dissolve the phosphate, and the insoluble substance removed by filtering, the solutions obtained were always more or less turbid, and the degree of turbidity was found to be related to the amount of phosphate present and to the shape of loaf produced. on further investigation it was found that the turbidity was due to the fact that the concentration of acid and salts which make gluten coherent, also dissolve some of it, and gluten like other colloids gives a turbid solution. it was also found that the amount of gluten dissolved, and consequently the degree of turbidity, is related to the shape of the loaf which the flour will produce. 73 now it is quite easy to measure the degree of turbidity of a solution by pouring the solution into a glass vessel below which a small electric lamp is placed, and noting the depth of the liquid through which the filament of the lamp can just be seen. the turbidities were, however, so slight that it was found necessary to increase them by adding a little iodine solution which gives a brown milkiness with solutions of gluten, the degree of milkiness depending on the amount of gluten in the solution. in this way a method was devised which is rapid, easy, and can be carried out with so little wheat that the produce of one ear is amply sufficient. it can therefore be used by the plant breeder for picking out from the progeny of his crosses those individual plants which are likely to give shapely loaves. the method is as follows: an ear of wheat is rubbed out and ground to powder in a small mill. one gram of this powder, or of flour if that is to be tested, is weighed out and put into a small bottle. to it is added 20 c.c. of water. the bottle is then shaken for one hour. at the end of this time the contents are poured onto a filter. to 15 c.c. of the solution 1? c.c. of a weak solution of iodine is added, and after standing for half an hour the turbidity test is applied. working in this way it is possible to see through only 10 c.m. of the solution thus obtained from such a wheat as red fife, as compared with 25 c.m. in the case of rivet. 74 other wheats yield solutions of intermediate opacity. this method is now being tested in connection with the cambridge wheat breeding experiments.