To determine if CO_{2} is given off during luminescence it is necessary to work with fairly pure luminous materials, obtained from luminous organisms. It is impossible to use the living organisms themselves as the CO_{2} continually respired becomes a very disturbing factor. From _Cypridina_, a small crustacean, two materials soluble in water may be prepared (_luciferin_ and _luciferase_), which will give a brilliant luminescence on mixing. It is possible to determine the H-ion concentration of the two solutions separately and of the mixture of the two after the luminescence has occurred.

If CO_{2} is produced during luminescence the H-ion concentration of the luminous solution should increase. Measurements made electrometrically with the hydrogen electrode have failed to demonstrate any increase in acidity. The PH of both solutions and of a mixture of the two is 9.04.

This would indicate that CO_{2} is not produced. As both luminous solutions contain proteins and the luminous substances themselves are probably proteins, which have a high buffer value, a method of this kind is none too sensitive. However, we can definitely state that not enough CO_{2} is produced to be detected and that this may be due to the buffer action of the luminous substances themselves. After all, unless luminescence is connected with respiration, we should hardly expect CO_{2} to be produced.

Another method of testing CO_{2} production is to measure the amount of heat produced during luminescence. Substances burned during respiration give off considerable heat, one gram of glucose to CO_{2} and H_{2}O, as much as 4000 calories. We have seen in Chapter III that no infra-red radiation is produced in the light of the firefly. This does not mean, however, that no heat is produced by the reaction which produces the luminescence. A temperature change of a few thousandths or hundredths of a degree would evolve no measurable radiation. Coblentz (1912) first studied the problem of heat production in the firefly, using a thermocouple as the measuring instrument. He came to the conclusion that the temperature of the insect was slightly lower than the temperature of the air and that the luminous segments were slightly hotter than the non-luminous segments, whereas a dead firefly is of the same temperature as its surroundings. No definite increase or decrease in temperature could be established during the flash of the firefly. However, further work on the firefly is much to be desired.

The use of a living animal for such measurements introduces a possible source of error in that any contraction of the muscles of the animal will produce heat which may add to an increase or mask a decrease of temperature during luminescence. Utilization of extracts of luminous animals containing the luciferin and luciferase mentioned above avoids the complications due to muscular contraction. By bringing the solutions of luciferin and luciferase to the same temperature and then mixing them one can measure any increase or decrease of temperature which occurs during the luminescence which results from mixing. We can thus gain some idea of the heat of oxidation of luciferin.

As a determination of heat production is of considerable interest the method will be given in some detail. Although the experiment sounds very simple, it is actually somewhat difficult to carry out. The attainment of temperature equilibrium between two solutions is very slow when one wishes to obtain them to within 0.001 C. of the same temperature. After many attempts, the following arrangement of apparatus (Fig. 33) was found most satisfactory. About 10 c.c. luciferin solution was placed in the inner tube (_D_) of a special non-silvered thermos bottle (_A_).

About 1 c.c. of luciferase solution was placed in a very thin-walled gla.s.s tube (_E_) which was immersed in the luciferin solution and connected with a small motor so that it could be slowly but constantly rotated, thus stirring the solutions. Thermocouples (_L_ and _M_) of advance (.008 in)--copper (No. 30, _B_ and _S_, enamel insulated) wire were paraffined and placed in each tube and the copper wires connected through a copper double throw switch (_C_) with a Leeds and Northrup d'Arsonval wall galvanometer (No. 34637, silver strip suspension) of 35 ohms resistance and 310 megohms sensitivity. The constant temperature junctions (_N_) were placed in a large Dewar flask (_B_) filled with water at approximately the same temperature as the luciferin solution.

One mm. galvanometer scale division represented 0.003 C. and the division readings could be estimated to tenths. By means of a gla.s.s rod (_F_) placed in the tube containing luciferase solution, this tube could be broken and the luciferase and luciferin solution mixed.

[Ill.u.s.tration: FIG. 33.--Apparatus for determining heat production during luminescence of luciferin. A, special thermos tube. B, Dewar flask for constant temperature junctions. C, double throw switch. D, tube containing luciferin solution. E, tube containing luciferase solution. F, gla.s.s rod for breaking E. G, rubber stopper with groove, K, for pulley cord. H, cork closing thermos tube. J, bra.s.s sleeve in H allowing rotation of E. L, thermojunction in luciferase solution. M, thermojunction in luciferin solution. N, constant temperature junctions.]

[Ill.u.s.tration: FIG. 34.--Curve showing temperature change when two tubes containing water at the same temperature are mixed. 0.1 galvanometer scale division = 0.003 C. Dots represent readings of thermocouple in tube D; crosses readings of thermocouple in tube E.]

It was found that even after the luciferase and luciferin solutions came to the same temperature within the thermos bottle, this was not necessarily the same as that of the room and a slow rise or fall occurred as indicated by a slow drift of the galvanometer coil. Readings of each thermocouple on the galvanometer scale were therefore taken at one-minute intervals for some time before and after mixing the luciferin and luciferase solutions and plotted as curves. Control experiments were also carried out in exactly the same manner as the luciferin-luciferase experiments, but water was placed in the two tubes instead of luciferin and luciferase. Figs. 34 and 35 give typical experiments with water and with luminescent solutions, respectively.

[Ill.u.s.tration: FIG. 35.--Curve showing temperature change when luciferin and luciferase solutions at the same temperature are mixed. 0.1 galvanometer scale division = 0.003 C. Dots represent readings of thermocouple in luciferin solution; crosses, readings of thermocouple in luciferase solution.]

With both control (water) and luciferin experiments there was a slight rise in temperature on mixing the liquids in the two tubes. The average rise of five control (water) experiments was .0054 C. and the average rise of five luciferin experiments was .0048 C.

The average rise in temperature is no doubt due to heat from friction in mixing of the liquids and breaking of the gla.s.s tube. The difference in the average rise of control and of luciferin experiments is so small (.0006 C.) as to have little significance. We may therefore conclude that if any temperature change occurs during the luminescent reaction it is certainly less than 0.001 C. and probably less than 0.0005 C., too small to be measured by this method.

To prepare the luciferin solution, two grams of dried _Cypridina_ were dissolved in 20 c.c. hot water and 10 c.c. of this 10 per cent. solution was used in the thermos bottle in the above experiments. If we a.s.sume that 1 per cent. of the dried _Cypridina_ is luciferin, 0.01 gram of luciferin on oxidation was not able to raise the temperature of the 10 c.c. (in reality 11 c.c., since 1 c.c. luciferase solution was mixed with the 10 c.c. luciferin solution) .001 C. This means that 1 gram luciferin liberates _at least less_ than 10 calories during the luminescence accompanying oxidation.

Since 1 gram glucose liberates 4000 calories on complete oxidation to CO_{2} and H_{2}O, it will be seen that the oxidation of luciferin is a very different type of reaction from the oxidation of glucose. As we shall see, it is probably similar to the oxidation of reduced haemoglobin or the oxidation of leuco methylene-blue to methylene blue. According to Barcroft and Hill (1910), 1.85 calories are produced per gram of haemoglobin oxidized. I have been unable to find figures for the heat exchange during oxidation of leuco-dyes, but it is no doubt also small.

Since luciferin evolves no measurable amount of heat on oxidation, we have very good evidence in support of that obtained by electrometric measurements of H-ion concentration, that no carbon dioxide is produced during luminescence of luminous animals.

In most animal cells it is perfectly clear that luminescence does not accompany respiration, since respiration is a continuous process, whereas light is only produced on stimulation. It is true that on stimulation respiration is accelerated, and we might suppose that luminescence is an accompaniment of accelerated respiratory oxidations; but this is not the case, for in luminous animals a rise in temperature of ten degrees centigrade will accelerate the respiratory oxidations 250 per cent. without necessarily causing the production of light.

In fungi and bacteria, on the other hand, which continually emit light, it is quite natural to suppose that the light is an accompaniment of respiration, just as we know the heat of these forms to be. This view was accepted by such of the earlier workers as Fabre in 1855, who found that luminous portions of a mushroom, _Agaricus olearius_, gave off more CO_{2} (4.41 c.c. CO_{2} per gram in 36 hours at 12 C.) than non-luminous portions (2.88 c.c. CO_{2} per gram in 36 hours at 12 C.).

This experiment has never been repeated and there are many reasons besides luminescence why one piece of fungus might have a more rapid respiratory rate than another piece. It is not true that rapidly respiring plant tissues, such as germinating seeds or the spadix of _Araceae_, are luminous, although they produce considerable heat.

On the other hand, it is very easy to prove that luminescence, even in bacteria, is not connected with respiration. Thus, Beijerinck (1889 _c_) found that of several species of luminous bacteria studied by him, one, _Bacterium phosph.o.r.escens_, was a facultative anaerobe and would grow, _i.e._, multiply, but not luminesce in the absence of oxygen. Some forms, ordinarily producing light, will grow, but fail to luminesce at high temperatures. Beijerinck (1915) has recently found that these individuals may, by continued cultivation at high temperatures, form non-luminous strains which fail to luminesce when again brought into lower temperatures, favorable for luminescence. These non-luminous mutants occasionally give rise to atavistic brilliantly luminous forms.

Beijerinck also finds that after exposure of _Photobacter splendidum_ to ultra-violet or strong sunlight, radium or mesothorium rays, luminescence continues but no growth occurs. There is thus ample evidence that growth and respiration are properties quite distinct and separable from luminescence. Indeed, respiration increases continuously up to a relatively high maximum whereas luminescence falls off rapidly above a relatively low optimum. McKenney (1902) found also that _Bacillus phosph.o.r.escens_ could grow rapidly in 0.5 per cent. ether without producing light.

Luminescence has been compared in bacteria to pigment formation, as rather definite cultural conditions are necessary for realization of both chromogenic and photogenic function. Some pigment-formers, as _Bacillus pyocyaneus_, which produces a water-soluble green pigment, remain colorless under anaerobic conditions. A colorless chromogen is formed, which oxidizes to the green pigment in the air. If this colorless chromogen produced light during its oxidation as well as green pigment, we would have a case of both chromogenic and photogenic function combined in one species of bacterium. Luminescence involves something more than respiration, an oxidation of a very definite and particular kind.

Since Spallanzani's observation that the luminous material of medusae could be dried, and upon moistening would again give light, many confirmatory observations have been made on other forms. _Pyrosoma_, _Pholas_, _Phyllirrhoe_, fireflies, _Pyrophorus_, copepods, ostracods, pennatulids, fungi, and bacteria can all be dessicated and the photogenic material preserved for a greater or less time. In a dessicator filled with CaCl_{2}, dried luminous bacteria lose, after a few months, their power to give light on being moistened. On the other hand, ostracods and copepods will still luminesce after years of dessication. The luminous material in the latter case appears capable of indefinite preservation, but it is possible that the quick loss of photogenic power with dried luminous bacteria is merely an indication that they contain very little photogenic substance and that the dried ostracods would also in time lose their power to luminesce. It is certainly a fact that the amount of luminous material in a single gland cell of an ostracod is vastly greater than that in the same ma.s.s of bacterial colony.

When the dried powdered luminous material of an ostracod is sprinkled over the surface of water, it goes into solution and leaves luminous diffusion and convection trails plainly visible in the water. Many luminous marine forms give off a phosph.o.r.escent slime when they are handled, which adheres to the fingers. It is not surprising that this luminous matter should have early received a name. In 1872, Phipson called it _noctilucin_ and described some of its properties. He regarded the luminous matter which can be sc.r.a.ped from dead fish (luminous bacteria) and the mucous secretion of _Scolopendra electrica_ or the luminous matter of the glowworm to be this material, noctilucin, which, "in moist condition, takes up oxygen and gives off CO_{2} and when dry appears like mucin." Phipson says that it forms an oily layer over the seas in summer (he probably refers to ma.s.ses of dinoflagellates), is liquid at ordinary temperatures and less dense than water, smells a little like caprylic acid, is insoluble in water but miscible with it, insoluble in alcohol and ether, dissolves with decomposition in mineral acids and alkalies and contains no phosphorus. We can see from this description that the word "noctilucin" does not indicate a chemical individual, but it is the earliest attempt to definitely designate the luminous substance.

The idea of a definite substance oxidizing and causing the light has been upheld by a number of investigators, and many years later Molisch called this substance the _photogen_. He contrasts the "photogen theory"

with certain other views of light production, which may be spoken of as "vital theories," notably those of Pfluger (1875), who looked upon luminescence as a sign of intense respiration, and of Beijerinck (1915), who regarded the light as an accompaniment of the formation of living matter from peptone.

Fortunately biological science has advanced beyond the stage where a living process can be explained by calling it a vital process, and we must fall back upon the idea of a photogen oxidizing with light production. Indeed, it is now possible to go much further than this and describe the properties of the photogen, but we must not lose sight of the fact that it was recognized very early in the history of Bioluminescence, that water, oxygen, and a photogenic substance were necessary for light production.

A very great advance in our knowledge of the chemistry of the problem was made by Dubois in 1885. He showed that if one dips the luminous organ of _Pyrophorus_ in hot water, the light disappears and will not return again. Also if one grinds up a luminous organ the ma.s.s will glow for some time but the light soon disappears. If one brings the previously heated organ in contact with the unheated triturated organ it will again give off light. Later, Dubois showed that the same experiment could be performed with the luminous tissues of _Pholas dactylus_. A hot-water extract of the luminous tissue, and a cold-water extract of the luminous tissue, allowed to stand until the light disappears, will again produce light if mixed together. Dubois (1887 _b_) advanced the theory that in the hot-water extract there is a substance, luciferin, not destroyed by heating, which oxidizes with light production in the presence of an enzyme, luciferase, which is destroyed on heating. The luciferase is present together with luciferin in the cold-water extract, but the luciferin is soon oxidized and luciferase alone remains. Mixing a solution of luciferin and luciferase always results in light production until the luciferin is again oxidized. Similar substances have been found by me in the American fireflies, _Photinus_ and _Photuris_, the j.a.panese firefly, _Luciola_, and in the ostracod crustacean, _Cypridina hilgendorfii_. Crozier[6] reports that they exist also in _Ptychodera_, a balanoglossid. I have been unable to demonstrate their existence in luminous bacteria; in the annelid, _Chaetopterus_; the pennatulids, _Cavernularia_ and _Pennatula_; the squid, _Watasenia_; and the fish, _Monocentris j.a.ponica_. E. B. Harvey (1917) could not demonstrate them in _Noctiluca_. There are several reasons why the existence of such bodies might be difficult to demonstrate, but these reasons cannot be considered here. We thus see that the photogen is in reality of dual nature, that two substances are necessary for light production and that they may be very readily separated because of difference in resistance to heating. In this respect Bioluminescence is similar to some other biological processes, notably to certain immune reactions and to certain enzyme actions.

[6] Private communication.

Thus, for the haemolysis of foreign red blood corpuscles, a specific immune body (_amboceptor or substance sensibilatrice_) not destroyed by moderate heating, and a thermolabile complement (_alexin_) are necessary.

For the alcoholic fermentation of glucose by the zymase of yeast juice two substances are also necessary. The zymase is made up of a heat resistant, dialyzing component, the co-enzyme, and a non-dialyzing substance, destroyed on boiling, the enzyme proper. Both must be present for alcoholic fermentation of glucose to proceed and the two may be separated by dialysis or by their difference in resistance to heating.

Several other characteristics of living cells are known to depend on the joint action of two substances, one thermolabile, the other thermostable. The reducing action of tissues, according to Bach, requires a reducing enzyme proper or perhydridase and some easily oxidizable substance, such as an aldehyde. The aldehyde has been spoken of as the co-enzyme.

Because of the necessity of thermostable and thermolabile substances for light production in luminous animals and because I was unable to oxidize the thermostable material of _Cypridina_ with such oxidizing agents as KMnO_{4}, H_{2}O_{2}, blood and H_{2}O_{2}, BaO_{2}, etc., I called the heat resistant substance of _Cypridina_, "_photophelein_" (from _phos_, light and _opheleo_, to a.s.sist), comparable to co-zymase, and the heat sensitive substance of _Cypridina_, "photogenin" (from _phos_, light and _gennao_, to produce), comparable to the zymase proper of yeast. In mode of preparation and properties, the photophelein of _Cypridina_ was also comparable to the luciferin of _Pholas_ and the _photogenin_ of _Cypridina_ to the luciferase of _Pholas_. I also regarded photogenin as the source of the light (hence the name), because a solution of _Cypridina_ photogenin (=_Pholas_ luciferase) will give light on mixing with crystals of salt and other substances which could not possibly be oxidized. I later found, however, that this result was due to the fact that the photogenin solution contained some of the thermostable substance (luciferin) bound (combined or adsorbed), and that this was freed by the salt crystals and oxidized with light production. I have consequently abandoned the view that the system of substances concerned in light production is similar to the zymase--co-zymase system of yeast--and have adopted Dubois' term, luciferase (=_photogenin_) for the thermolabile material, and luciferin (=_photophelein_) for the thermostable material.

The luciferin of _Cypridina_ differs from that of _Pholas_ in that it will not oxidize with light production with any oxidizing agents that I have tried, and will give no light with luciferase from _Pholas_. It does, however, oxidize spontaneously in solution, although no light accompanies this oxidation.

I believe that for accuracy and definiteness we must designate the luciferins and luciferases from different animals by prefixing the generic name of the animal and speak of _Pholas_ luciferin, _Cypridina_ luciferase, _Pyrophorus_ luciferase, etc. In extracts of many non-luminous animals Dubois has found oxidizing agents which can oxidize _Pholas_ luciferin with light production and I have confirmed this for _Pholas_, but I have not found any such substances in non-luminous animals which will oxidize _Cypridina_ luciferin with light production.

I have found in extracts of non-luminous animals substances which will liberate the bound luciferin in a concentrated _Cypridina_ luciferase solution. The luciferin can then be oxidized by the luciferase and light appears. Their effect is similar to that of salt crystals and I suggest that they be called _photopheleins_, substances that a.s.sist in the luciferin-luciferase reaction by liberating bound luciferin. One of the best ways of freeing a solution of luciferase from bound luciferin is to shake with chloroform. We can then do away with the disturbing effects of bound luciferin.

It is obvious that luciferin must be formed from some precursor in the cell and following the usual biochemical terminology, Dubois has called it _proluciferin_ or _preluciferin_, and believes that it is converted into luciferin by an enzyme co-luciferase. The experiments to prove the existence of proluciferin were first made by Dubois on _Pholas_ in 1907 and have since been amplified (1917 _a_; 1918 _a_ and _b_).

In order to understand these experiments it must be borne in mind that Dubois prepares luciferin from _Pholas_ in three ways: (1) By precipitating the viscid luminous fluid from the siphons with 95 alcohol and dissolving the precipitate in water (1901_a_, 1907). (2) By extracting the luminous organs with 90 alcohol in a closed vessel for twelve hours and filtering (1896). (3) By heating the viscid luminous fluid to 70 C. Apparently _Pholas_ luciferin is sparingly soluble in alcohol as it can be obtained either in an alcoholic extract (method 2) or by precipitation with alcohol (method 1). Proluciferin (called _preluciferine_ in a later paper, 1917 _a_, 1918 _a_), is prepared by methods 1 or 2 except that fatigued siphons, from which luciferin has been removed by washing, are used (1907, 1917 _a_, 1918 _a_).

Preluciferin can also be obtained on boiling an extract of the luminous organ of _Pholas_ because luciferin (at 70), luciferase (at 60) and a co-luciferase are all destroyed below the boiling point (1917 _a_).

Co-luciferase is prepared (1) by heating a luciferase solution to 65 (1917 _a_) or (2) by extracting with water portions of the siphon of _Pholas_ which have previously been macerated and well extracted with alcohol (1918 _a_). Long-continued treatment with alcohol apparently destroys the luciferase without affecting the co-luciferase. On mixing a solution of preluciferin with one of co-luciferase and allowing them to stand for 8-10 hours, luciferase is formed and can be recognized by the fact that it will give light with a crystal of KMnO_{4}. Preluciferine does not do this.

Recently Dubois (1918 _a_) states that preluciferine is nothing but taurine and that taurine occurs in large quant.i.ties in _Pholas_ and is transformed into luciferine by the action of co-luciferase. Not only taurine, but also Byla's peptone, egg lecithin, and esculin can be converted into luciferine by co-luciferase, and since esculin, a glucoside, is so transformed, Dubois believes this proves that co-luciferase belongs to the hydrolases. Indeed, it proves too much.

Luciferin must have an extraordinary chemical structure if it can be formed by hydrolysis of such diverse compounds as peptone, lecithin, esculin and taurine. A glance at the structural formula of esculin and taurine is sufficient to emphasize the diverse nature of these two substances.

[Ill.u.s.tration: Taurine]

[Ill.u.s.tration: Esculin]

I believe that in these experiments Dubois has been working with an oxidation product of luciferin, what I have called _oxyluciferin_, rather than a pro-substance. The mode of preparation of _Pholas_ preluciferin and _Pholas_ co-luciferase is such as could be used in the preparation of _Cypridina_ oxyluciferin, and it seems more logical to look for the presence of _Pholas_ oxyluciferin in one or both of Dubois'

extracts rather than believe that luciferin can be formed from both taurine and esculin. When the co-luciferase solution stands with the preluciferin solution we would in reality have not the formation of luciferin from preluciferin, but the formation of luciferin from oxyluciferin, by some reducing agent in the mixture. Indeed, in a very recent paper Dubois (1919 _c_) takes the view that his co-luciferase is a reducing enzyme which forms luciferin by reduction (presumably from oxidized luciferin) and no mention is made of preluciferin.

It is, of course, obvious that when luciferin oxidizes, some oxidation products must be formed. Most observers have a.s.sumed the oxidation products of luciferin to be relatively simple and to represent a rather complete breaking down of the luciferin molecule. Carbon dioxide was mentioned by Phipson (1872) as being formed. We have just seen that no carbon dioxide is formed during the oxidation of _Cypridina_ luciferin and there is evidence that no fundamental change at all occurs. It is for this reason that I have called the oxidation product of luciferin _oxyluciferin_.[7] As we shall later see, the change luciferin oxyluciferin is to be compared to the oxidation of colorless dyes (leuco-compounds) to the colored dye. The chemical properties of oxyluciferin are similar to those of luciferin and the oxyluciferin can be readily reduced to luciferin again.

[7] It is unfortunate that Dubois (1918 b) has used the term oxyluciferine in a quite different sense from the present use. He regards oxyluciferine as a substance still capable of giving light by autooxidation, and represents the steps in luminescence as follows:

"Co-luciferase + preluciferine = luciferine.

Luciferase + luciferine = oxyluciferine.

Oxyluciferine + oxygene = lumiere."

I should represent them as follows:

Luciferin + oxygen ? oxyluciferin.

The reaction proceeds to right with light production only in presence of luciferase.

Finally, we have the fluorescent substance of _Pyrophorus_ and fireflies, which Dubois first called _pyrophorin_, but later, adopting McDermott's terminology, speaks of as _luciferesceine_. This Dubois regards as a substance intensifying the light and modifying its color by changing invisible into visible rays. As we have seen, this theory, while attractive, will not stand the test of critical examination.