SPECTROSCOPY IN BIOLOGY

E. V. Shpol’skii

To anyone who observes the development of modern natural science, a healthy tendency that has recently become apparent is clear: to attack complex problems from different sides, to approach their solution through the combined efforts of representatives of various specialties. Aristotle’s physics was a science of nature as a whole. Gradually, however, the accumulation of an enormous number of facts and the deepening and complication of research methods led to ever narrower specialization. In the second half of the nineteenth century, only the genius of Helmholtz was capable of simultaneously embracing mathematics, mechanics, physics, physiology, medicine, and the theory of art (music and painting). In the first thirty years of the twentieth century we see such fractional specialization that even representatives of the same science—for example theoretical physicists and experimentalists—sometimes cease to understand one another. At the same time, the impracticability becomes clear of attempts to solve the most complex problems of natural science, especially biology, in which the most diverse factors—physical, chemical, and purely biological—are intricately intertwined, by the efforts of only “narrow” specialists.

If, however, the ramification and specialization of modern natural science is so great that one person is not in a position to be master of several disciplines and to command all existing methods of research, then the only way out is for several specialists to unite for the joint solution of a single problem. And here a new feature reveals itself: works appear with three, four, five authors—representatives of different specialties.

It is well known how refined the modern experimental methods of physics are. Electron microscopy, which grew out of profound theoretical constructions that at first had no relation whatever to the tasks of experimental technique, makes it possible to see individual large molecules; the optical spectrograph and the spectral monochromator, in combination with the photoelement, make it possible to...

in short intervals of time, measured in minutes, to carry out complex analyses which previously required days or weeks, and sometimes were altogether impossible; radioactive indicators, Geiger counters, and ionization chambers make it possible to observe processes in “weightless” quantities of substance, etc., etc. It would be at least unreasonable not to use all these powerful means for solving the profound and fascinating problems of living matter, or the noble tasks of healing the sick.

Among the various physical methods of investigation, optical methods have in recent times found the widest application in biology and biochemistry,* and among the optical methods spectroscopy opens up the most varied possibilities. All kinds of spectra and all methods of exciting them are used.

A complete and systematic review of the diverse applications of spectroscopy in biology could serve as the subject of a large monograph. The task of the present article is merely to give illustrations, using a few examples, of the application of absorption spectra in biology.

I

Let us first dwell briefly on certain general questions concerning the absorption spectra of complex organic compounds.

It has long been known that the absorption spectrum of a complex molecule is determined mainly not by the whole molecule as such, but by some definite part of it or by several parts. Such groups of atoms, on which the position and magnitude of the absorption maxima chiefly depend, are called chromophore groups. Their existence was first established above all in dyes, i.e., in substances absorbing in the visible part of the spectrum—whence the name itself derives. The same, however, also occurs in substances absorbing in the infrared and ultraviolet parts of the spectrum. At the same time, as might be expected, the absorption of chromophores depends not only on the nature of the atoms entering into the group, but also on the nature and distribution of the bonds.

In the infrared part of the spectrum the absorption of chromophores is determined by the frequencies of vibrations (valence vibrations are meant)

* A large review by J. Loofbourow (about 10 printed sheets and an index of the original literature containing 1203 works), “Borderline Problems of Biology and Physics,” published in Reviews of Modern Physics, contains the following sections: I. Isotopes as biological indicators. II. Study of biochemical substances, tissues, etc., by means of X-rays. III. Application of infrared, Raman, visible, and ultraviolet spectra to biological problems. IV. Application of ultracentrifugation to biology and biochemistry. V. The latest development of microscopy (ultraviolet and electron microscopy). VI. Various other methods.

and rotation. Characteristic frequencies are found with only small changes in different molecules containing the given chromophoric group. Examples may be the chromophores¹

\[ \begin{array}{cc} \begin{matrix} \backslash\\[-6pt] \mathrm{C}{-}\mathrm{H}\\[-6pt] / \end{matrix} & 2910\ \mathrm{cm}^{-1} \\[8pt] \begin{matrix} \backslash\\[-6pt] \mathrm{C}{=}\mathrm{C}\\[-6pt] / \end{matrix} & 1630\ \mathrm{cm}^{-1} \\[8pt] \begin{matrix} \backslash\\[-6pt] {-}\mathrm{C}{-}\mathrm{C}{-}\\[-6pt] / \end{matrix} & 990\ \mathrm{cm}^{-1} \\[8pt] \begin{matrix} \backslash\\[-6pt] \mathrm{C}{=}\mathrm{O}\\[-6pt] / \end{matrix} & 1722\ \mathrm{cm}^{-1} \\[8pt] {\equiv}\mathrm{C}{-}\mathrm{O}{-} & 1034\ \mathrm{cm}^{-1} \end{array} \]

In the visible and ultraviolet parts of the spectrum, absorption is determined by electronic transitions. Here too, in many cases it is possible to establish the chromophoric group, as was first shown on extensive material by Anri’s systematic investigations.²

Thus, for example, the presence of the carbonyl group

\[ \begin{matrix} \backslash\\[-6pt] \mathrm{C}{=}\mathrm{O}\\[-6pt] / \end{matrix} \]

causes the appearance in the ultraviolet spectrum of a very definite, although comparatively weak, band with a maximum near 2700—2800 Å. The chromophoric character of this group is seen from the following table, which gives the absorption maxima and molar extinction coefficients*) for various ketones.

Another example of chromophores may be the ethylene group

\[ \begin{matrix} \backslash\\[-6pt] \mathrm{C}{=}\mathrm{C}\\[-6pt] / \end{matrix} \]

The properties of this chromophore are very interesting. In ethylene the absorption band lies in the far ultraviolet region of the spectrum, with a maximum at 1645 Å. When the hydrogen atoms are replaced by radicals

*) The intensity of absorption is usually characterized by the so-called molar extinction coefficient \(\varepsilon\), entering into the formula \(I = I_0 \cdot 10^{-\varepsilon c l}\), where \(I\) is the intensity of the light that has passed through a layer \(l\) cm thick of a substance whose molar concentration is equal to \(c\).

(alkyls), the band is regularly displaced toward the long-wavelength side, as is seen from the following examples³:

Benzene ring

$\mathrm{C_6H_6}$

also gives a very characteristic group of bands, which is preserved also in benzene derivatives. In this group of bands one can recognize the superposition, upon an electronic transition, of definite vibrational states.

The chromophoric properties of various groups of atoms in the ultraviolet part of the spectrum are apparently due to the presence in these groups of weakly bound or free electrons. This is especially clearly seen in the case of the benzene ring. It is known that the six electrons of the benzene ring (the so-called $\pi$-electrons) behave as free electrons, causing, among other things, the very large value of the diamagnetic susceptibility of the benzene molecule. It may also be noted that the lowest ionization potential and the chemical behavior of the molecule are essentially determined by the same unbound electrons of the chromophores, whence it is evident how essential the study of ultraviolet absorption spectra is for chemistry.

If a molecule contains two or several chromophoric groups, the result proves to be different depending on the nature of the bond between the chromophores and on their position in the molecule. In some cases, despite the presence of two chromophores, the absorption spectrum shows absorption characteristic only

of one of them. Thus, for example, the molecule of benzaldehyde $\mathrm{C_6H_5\!-\!CHO}$ contains two chromophores: carbonyl $\mathrm{>CO}$ and the phenyl group $\mathrm{C_6H_5}$ (benzene ring). Nevertheless, benzaldehyde shows only an absorption spectrum located between 2700 and 2400 Å.

and characteristic of the benzene ring; the presence of the carbonyl group has no effect on the absorption spectrum, since in the region of 2800 Å no absorption is observed.

On the other hand, very often the presence of two or several chromophores is clearly manifested in the absorption spectrum, and in different ways. In those cases where carbon atoms with a double bond \(C=C\), or at least two simple bonds \(C—C—C\), are situated between the chromophores, the absorption spectrum is additively composed of the absorption spectra of the individual chromophores. A different picture is obtained in the case where chromophores with a \(C=C\) bond are separated by only one simple \(C—C\) bond. For example, two ethylenic chromophores may be linked as follows:

\[ >C=C-C=C< . \]

An unsaturated bond of this kind is called conjugated. Chromophores connected by conjugated bonds may form long chains

\[ >C=C-C=C-C=C\ldots-C=C< \]

They may also form closed rings. In particular, the benzene ring may be regarded as a sequence of conjugated bonds closed upon itself.

The appearance of conjugated bonds leads to extremely sharp changes in the absorption spectrum. With a successive increase in the number of such bonds, the entire spectrum shifts toward the long-wave side and, when a sufficient number of them has accumulated, moves from the far ultraviolet into the visible region. At the same time, the intensity of absorption, measured by the molar extinction coefficient \(\varepsilon\), increases greatly.

Let us consider in somewhat more detail the properties of linear molecules with conjugated bonds that are important for biology. In all these molecules the chief chromophore is the polyene chain

\[ —(HC=CH)_n—, \]

to which various groups may be attached at the ends. For example,\(^4\) series of polyene aldehydes have been studied:

\[ CH_3—(HC=CH)_n—CHO \qquad n=1,\ 2,\ 3, \]

polyene carboxylic acids:

\[ CH_3—(HC=CH)_n—COOH \qquad n=1,\ 2,\ 3,\ 4, \]

diphenyl-polyenes

\[ C_6H_5—(HC=CH)_n—C_6H_5 \qquad n=1,\ 2,\ 3,\ 4,\ 5,\ 6,\ 7. \]

A typical example of such molecules is provided by the last series of diphenyl-polyenes.

The absorption spectra of the molecules of this series are shown in Fig. 1. The shift of the spectrum, described above, with an increase in the number of conjugated bonds is seen here very clearly.

Fig. 1. Absorption spectra of diphenyl-polyenes (after Kuhn, Hausser, and Smakula).

How regular this shift is may be judged from the following interesting facts. If the number of ethylene groups is plotted along the abscissa axis, and the molar extinction coefficient \(\varepsilon_{\max}\) along the ordinate axis, a straight line is obtained (Fig. 2). A linear dependence is also obtained if the reciprocal of the square root of the reduced mass of the molecule is plotted along the abscissa axis (Fig. 3), and the frequency of the principal maximum along the ordinate axis.

Taking into account the known relation between the vibration frequency of a linear oscillator and its reduced mass, this result may be interpreted as indicating that the principal band arises as a result of the superposition of vibrations of the polyene chain on the electronic transition.

From the biological point of view, of great interest are the so-called carotenoids, whose molecules also have a chain character. The well-known orange-yellow pigment occurring in plant leaves and playing some role, as yet not clarified, in the process of photosynthesis is one example. In fact this pigment, usually called simply carotene, is a mixture of three chemically isomeric substances: \(\alpha\)-, \(\beta\)-, and \(\gamma\)-carotene. The pigment that colors tomatoes—lycopene—also belongs to the carotenoids. The absorption spectra of various carotenoids are shown in Fig. 4. As can be seen, they exhibit a great similarity to the spectra of the diphenyl-polyenes just considered. The reason for this is clear. The structure of carotenoids is characterized by a long chain composed of dehydrogenated isopre-

SPECTROSCOPY IN BIOLOGY

of new units

\[ -\mathrm{CH}=\mathrm{C}-\mathrm{CH}=\mathrm{CH}- \]
\[ \phantom{-\mathrm{CH}=}\big| \qquad \mathrm{CH_3} \]

If such a unit is denoted by \(ip\), then the structure of \(\beta\)-carotene is represented in the form:

\[ \begin{array}{c} \text{benzene ring}-ip-ip-pi-pi-\text{benzene ring} \end{array} \]
\[ \text{with } \mathrm{CH_3} \text{ substituents at the ends.} \]

The formulas given show that the long chain characterizing the structure of \(\beta\)-carotene, in essence, as in all polyenes,

Fig. 2. Dependence of the molar extinction coefficient of diphenyl-polyenes on the number of ethylene groups.

Fig. 3. Dependence of the frequency of the principal absorption maximum of diphenyl-polyenes on the reduced mass of the molecule.

In Fig. 3:
\(\circ\) \(\mathrm{C_6H_5-(CH{=}CH)_n-C_6H_5}\)
\(\bullet\) \(\beta\)-carotene

Fig. 4. Absorption spectra of carotenoids (after Smakula).

a: \(\beta\)-carotene; \(\alpha\)-carotene.
b: taraxanthin (from dandelion); lutein (from yolk).

is formed by the accumulation of ethylene chromophores in a conjugated bond.

The following fact shows that the absorption of β-carotene is due precisely to ethylene linkages: if one calculates the reduced mass of the β-carotene molecule and plots on the graph of Fig. 3 the value for the reciprocal of the square root of it and the frequency of the principal absorption band, then the corresponding point lies on the straight line for diphenyl-polyenes.

This fact also has substantial practical significance. Indeed, it indicates that the absorption of light can be calculated from the number of certain bonds, just as is the case with molar refraction. If one takes into account that, in order to establish absorption, as little as \(0.1—0.5\) mg of substance often proves sufficient, then the importance of this method for solving complex structural problems of organic chemistry and biochemistry becomes evident.

The biological significance of the carotenoids is not limited to their role in the leaf of a plant. It is known that vitamin \(A\) is not found in plant food, but is produced in the animal organism from a certain substance serving as the starting material for its formation (provitamin). This provitamin is precisely β-carotene. This can already be concluded from the fact that the molecule of vitamin \(A\), in its structure, is basically simply one half of the symmetrical molecule of β-carotene.

\[ \begin{array}{c} \text{[benzene ring with } \mathrm{CH_3}\text{ substituent]}-\mathrm{ip}-\mathrm{ip}-\mathrm{CH_2OH} \qquad (\text{Vitamin } A) \end{array} \]

Hence follows the following picture of the formation of vitamin \(A\). In the animal organism β-carotene is split into two equal parts and, after the addition of water to each part, gives two molecules of vitamin \(A\)*). This scheme is confirmed by the following observation: if an animal is fed β-carotene, then in the absorption spectrum of the extract of its liver a band of vitamin \(A\) is found.

Investigation of the absorption spectrum of the pigment that colors the visual purple (rhodopsin) showed that it also belongs to the carotenoids and that there is a direct connection between the presence of vitamin \(A\) in the organism and the restoration of the visual purple. Hence follow a number of frequently practical conclusions concerning the role of vitamin \(A\) in night vision, etc.

The theory of the absorption of light by complex, in particular biologically important, molecules has recently attracted the steadily increasing attention of theoretical physicists. As early as Andri attempted to approach the calculation of absorption spectra of complex molecules with the aid of co-

) Attempts to carry out such a reaction in vitro* have not given a positive result. However, it apparently can proceed photochemically. This follows from the fact that in the absorption spectrum of β-carotene there is diffuse absorption in the region of the absorption maximum of vitamin \(A\).

disturbances based on the classical theory of dispersion, and obtained, in doing so, a number of encouraging results.

Of course, the adequate apparatus for the theory of molecules is provided not by classical mechanics, but by quantum mechanics. A number of quantum-mechanical works devoted to the absorption spectra of complex molecules (such, for example, as Sklar’s works⁶ on the theory of the electronic absorption spectra of benzene and its derivatives, Mulliken’s⁵ on the theory of the absorption of ethylene and its derivatives, etc.), which have appeared in recent times, is evidence of theorists’ interest in these important problems. However, the large number of particles entering into complex molecules permits one to think that purely classical considerations too may yield valuable results here (as is the case, for example, with the atomic nucleus).

II

Let us now turn to the consideration of individual examples of the application of spectroscopy in biology. One of the most striking examples is the history of the investigation of the antirachitic vitamin—vitamin $D$. It was known that improper calcium and phosphorus metabolism leads to the disease rickets, to which children, as well as young animals and birds, are susceptible. Experiments carried out with young rats showed that, with one-sided nutrition in the absence of vitamin $D$, the animals develop a severe disease, expressed in insufficient calcination of the bones. The disease could be cured by a corresponding diet, but it was also cured when the animals were irradiated with ultraviolet light. Evidently, the skin of the animal contains, in negligible quantities, a substance (the so-called provitamin $D$; its chemical composition is now known) which, upon irradiation, undergoes a photochemical reaction, producing vitamin $D$, and this latter is then carried by the flow of blood throughout the organism.

On further investigation it turned out that, for a cure, it was sufficient to irradiate the animals’ food with ultraviolet rays—the very same vitamin-free food that had caused the disease. From this it had to be concluded that this food too contains a provitamin which is photochemically converted into vitamin $D$. The American physiologist Hess, who was engaged in these investigations, advanced the hypothesis that this provitamin is cholesterol, present in the skin of rats as well as in their vitamin-free food. The basis for this, among other things, was the fact that cholesterol, when irradiated with ultraviolet light, undergoes a photochemical reaction: its transparency for wavelengths around 300 mμ increases somewhat after irradiation; consequently, it “fades.” Wishing to clarify this question completely, Hess turned to the Göttingen chemist Windaus, known for his work in the field of sterols. Being convinced that the provitamin $D$ was cholesterol,

Hess intended, with the aid of Windaus, to determine the chemical nature of vitamin \(D\) itself. However, the experiments carried out by Windaus did not confirm the identity of provitamin \(D\) with cholesterol. Then the physicist R. Pohl took part in resolving the problem; using a photoelectric spectrophotometer, he accurately measured the absorption spectrum of cholesterol before and after irradiation. The two spectra proved to be substantially different (Fig. 5): whereas three bands \(\alpha, \beta, \gamma\) were clearly superposed on the monotonically rising curve toward the short waves in the spectrum of unirradiated cholesterol, in the irradiated one they were absent. By subtracting curve \(b\) from curve \(a\), Pohl obtained a rough picture of these bands. In addition, he established that the transparency of irradiated cholesterol at \(\lambda = 280\ \mathrm{m\mu}\) is more than twice the transparency of unirradiated cholesterol. From this it could be concluded that about half of the cholesterol had undergone a photochemical transformation. But this contradicted the results of Windaus, who, on the basis of chemical experiments, asserted that no more than a few tenths of a percent could have undergone transformation. It remained to conclude that the previous identification of cholesterol with provitamin \(D\) was erroneous and that the bands obtained by Pohl by subtracting the curves belonged to another substance, present as an insignificant impurity in the cholesterol and actually constituting provitamin \(D\). Its absorption in that case had to be so strong that Pohl conditionally called it a “dye.” By means of tenfold distillation in a high vacuum, Windaus succeeded in enriching the cholesterol with this substance approximately twofold, after which Pohl was already able to obtain its spectrum more accurately.

Fig. 5. Absorption spectra of cholesterol: \(a\)—before irradiation, \(b\)—after irradiation. The dotted curve was obtained by subtracting from curve \(a\) curve \(b\) and represents the absorption spectrum of the substance admixed to the cholesterol that undergoes a photochemical transformation upon irradiation (after Pohl, Naturwiss. 15, 435, 1927, figs. 3 and 4).

On the basis of chemical data Windaus assumed that the sought provitamin was another sterol—namely ergosterol. Comparison of the spectra of both substances confirmed this assumption (Fig. 6). Thus the problem was solved quickly and definitely thanks to the application of spectroscopic methods*). According to this

*) Subsequent work revealed a considerably more complex picture. In particular, it was established that another substance (7-dehydrochole-

On this subject Windhaus, in one of his lectures, remarked that, probably, chemists, using their own methods, would have obtained the same result; but whereas, thanks to the application of physical methods and the participation of experimental physicists, the question was clarified within a few days, chemists and physiologists, for whom only experiments on animals could serve as indicators, would have required no less than a year for this!

III

Let us now consider certain spectral properties of two colored substances having the greatest importance for the preservation of life on earth: the coloring substance of the blood—hemoglobin—and the green coloring substance of the plant leaf—chlorophyll. It is highly remarkable that both of these substances have a very similar structure. Hemoglobin is a stoichiometric compound of a dye—hemin—with a protein substance—globin. It is very probable that chlorophyll also contains, along with the dye, a protein. However, at present this question cannot be considered finally resolved (see below). Since we are dealing with the colored, i.e. absorbing, parts of hemoglobin and chlorophyll, both of these pigments have one and the same chromophore—the so-called porphyrin ring. An element of the structure of this ring is the pyrrole nucleus

Fig. 7. Structure of porphin.

Fig. 6. On the left, the absorption spectrum of provitamin \(D\), mixed with a 1% alcoholic solution of cholesterin. On the right, the absorption spectrum of a 0.001% solution of ergosterin in alcohol. The slight difference between these spectra is due to the fact that the right-hand spectrum was obtained indirectly (by subtracting curves).

\[ \begin{matrix} \mathrm{HC}—\mathrm{CH}\\ \| \quad \|\\ \mathrm{HC}\quad \mathrm{CH}\\ \backslash\quad /\\ \mathrm{N}\\ \mathrm{H} \end{matrix} \]

In the molecule of porphin (this compound was synthesized by Fischer) there are four such nuclei, connected by bridges of methine groups
\[ \text{—CH—} \]
Porphin has the same absorption spectrum as ergosterin, and upon illumination also gives an antirachitically active product. See on this matter Morton’s monograph (literature II, A).

\(= \mathrm{CH}-\), form the ring shown in Fig. 7. In this ring there alternate 9 double and 9 single bonds, forming a closed sequence of conjugated bonds, similar to what occurs in benzene. In the schematic Fig. 8 this closed system, in two modifications, is outlined by a black line.

Fig. 8. Scheme of the structure of the porphyrin chromophore.

Porphin, like the linear conjugated chain in carotenoids, is a most important biological chromophore. Derivatives of porphin are called porphyrins. The complex compound of one of the porphyrins with iron is hemin; the complex compound of another

Fig. 9. Structure of protochlorophyll and chlorophyll \(a\).

porphyrin with magnesium forms protochlorophyll. From protochlorophyll, by the addition of two hydrogen atoms in the IV pyrrole nucleus (see Fig. 9), chlorophyll \(a\) arises; chlorophyll \(b\) differs from chlorophyll \(a\) only in that the group \(\mathrm{CH_3}\) marked with an asterisk in Fig. 9 is replaced by the group \(\mathrm{CHO}\). It is significant that in all these cases the porphyrin ring remains untouched.

Absorption spectra of various porphyrins, as well as of the parent substance porphin, have been studied by Stern and his collaborators. All these spectra have a very typical structure. Characteristic are four bands between 500 and 700 mμ and a very strong band in the blue-violet part of the spectrum. An interesting analysis of the absorption spectrum of porphin and its derivatives in connection with structural-chemical data has recently been made by E. Rabinowitch,^11 who in this way established the scheme of the energy levels of these important molecules. In Fig. 10 a scheme is given for porphin. As may be seen, the orange-yellow system of bands is due to transitions from the ground state \(x\) to various vibrational states of the higher-lying electronically excited state \(A\), whereas the blue-violet band arises in transitions to the zero vibrational state of the next electronic state \(B\).

Fig. 10. Scheme of the energy levels of the porphin molecule according to E. Rabinowitch.

Fig. 11. Absorption spectra of chlorophylls \(a\) and \(b\) (according to Scheibe).

In Fig. 11 the absorption spectra of both modifications of chlorophyll are shown.^12 As may be seen, besides the bands characteristic of porphin and porphyrins, there is here also a strong red band, of fundamental importance for photosynthesis. This band is characteristic neither of porphin nor of porphyrins. E. I. Rabinowitch drew attention to the fact that the appearance of this band is a consequence of the already mentioned addition of two extra hydrogen atoms in the IV pyrrole ring. The addition of two more hydrogen atoms

in the II pyrrole nucleus converts chlorophyll into bacteriochlorophyll and at the same time leads to the appearance of still another band, located in the infrared part of the spectrum with a maximum at 800 mμ. Thanks to the presence of this band, the bacteria prove capable of photosynthesis under the action of infrared rays (for details see 29).

The absorption spectrum of hemoglobin and its derivatives has been studied in an enormous number of works, chiefly for the practical purpose of determining the hemoglobin content in blood (see the literature in VIII). A detailed theoretical interpretation of this spectrum in connection with the structure of hemoglobin and with the absorption spectra of the corresponding porphyrins, so far as we know, has not yet been carried out. Nevertheless, the general structure of the spectrum, determined by the presence of the porphyrin skeleton, is preserved here as well.

Fig. 12. Absorption spectrum of oxyhemoglobin.

Fig. 12 gives the absorption spectrum of the oxidized form of hemoglobin—oxyhemoglobin (HbO₂). The bands α and β, situated in the yellow-green part of the spectrum, are usually used in practice and have maxima at 576 and 540 mμ. These maxima clearly belong to the porphyrin skeleton, since they exactly reproduce the II and III maxima of the absorption spectrum of protoporphyrin (576 and 538 mμ), from which hemin is derived. The γ band also belongs to the porphyrin skeleton and corresponds to the transition $X \to B$ (Fig. 10). On the other hand, the protoporphyrin band with maximum at 504 mμ is observed neither in oxyhemoglobin nor in reduced hemoglobin (Hb); instead of the red protoporphyrin band at 631 mμ, hemoglobin exhibits in the red region a band at 750 mμ. The doublet band γ′ ± HbO₂ is in some way connected with the presence of the iron atom, since it is not observed in protoporphyrin. Finally, the ultraviolet band at 275 mμ evidently belongs to the protein part of the oxyhemoglobin molecule—globin—since all cyclic amino acids have maxima near this wavelength.

IV

All these detailed data on the structure of the molecules and spectra relate to substances—chlorophyll and hemoglobin—which are studied in a form extracted from their natural state—in

solutions. It should not be forgotten, however, that both hemoglobin and chlorophyll occur in nature and perform their most important functions in special microscopic structural units: the former—in red blood corpuscles, the erythrocytes; the latter—in chloroplasts. There is no doubt that operations connected with extraction may alter the substances themselves. Therefore the question is quite appropriate: are the substances that we study in solutions identical with the substances in their natural state?

As regards hemoglobin, in the course of the development of biochemistry the answers to this question have varied. In the seventies of the last century a theory advanced by the greatest specialist in this field, Hoppe-Seyler, enjoyed popularity. According to this theory, hemoglobin in solution and hemoglobin inside the erythrocyte are two entirely different substances. In order not to confuse them, Hoppe-Seyler even proposed different names for them: arterin and phlebin—for the substances inside the erythrocyte; hemoglobin—for the substance in an extract or in the crystalline state. However, the facts to which Hoppe-Seyler referred in support of his theory later received another, more plausible explanation, and the theory was abandoned.

Unexpectedly, however, it has been revived quite recently, and precisely in connection with spectroscopic work. In the interval from 1930 to 1938 a number of papers appeared by the Canadian biochemist Adams and his collaborators ^13, ^14, ^15, who found that, whereas in a transparent solution of hemoglobin all three bands characteristic of its spectrum—α, β, and γ—are always observed, the γ band disappears when hemoglobin is inside the erythrocyte. This was a matter of the complete disappearance of the band, and not of its displacement into the invisible part of the spectrum, which Adams would have had to notice if it existed, since a quartz spectrograph and photography of the spectrum were used. Trying to explain this phenomenon by physical causes, Adams carried out a series of additional experiments; however, not one of them gave the desired explanation. Moreover, a special experiment showed that when pure hemoglobin crystallizes out of solution in the form of the finest non-settling suspension, the γ band is observed. It thus followed that the optical inhomogeneity of the system (a suspension of microcrystals in a colorless transparent medium) does not in itself affect the absorption spectrum when the suspended particles are known to be hemoglobin crystals. But in the case in which the suspension consists of erythrocytes, the band is absent. From this Adams drew, it would seem, an inevitable logical conclusion: the substance located inside the erythrocyte differs from hemoglobin in its chemical nature. It was Adams who put forward the hypothesis according to which, inside the erythrocyte, an unstable compound arises between hemoglobin and the protein substance of the erythrocyte (stromatin). In support of

Under this hypothesis he carried out experiments in which he supposedly succeeded in obtaining “in a test tube” the unstable compound hemoglobin–stromatin, in the spectrum of which the γ band is not observed.

Adams’s experiments for some time did not attract the attention of biochemists. However, in 1940 the well-known Cambridge biochemist Keilin[^18] published in Nature a large article in which he described his own observations, confirming Adams’s results insofar as they concerned the “disappearance” of the γ band. Figure 13 gives the curves of the short-wavelength part of the absorption spectrum of oxyhemoglobin, borrowed from the work of Keilin and Hartree. As can be seen, the γ band is completely absent in the absorption spectrum of erythrocytes. However, Keilin could not reproduce the hemoglobin–stromatin compound described by Adams. On the other hand, he showed that if an emulsion of a concentrated solution of pure hemoglobin in colorless paraffin oil is prepared, then in such an emulsion the γ band is likewise not observed, although after centrifugation the hemoglobin proves to be unchanged. On this basis Keilin spoke in favor of a physical explanation of the interesting phenomenon observed by Adams, emphasizing that, from the whole body of facts, no elementary explanation is possible here.

Fig. 13. Absorption spectrum of hemoglobin in solution and in erythrocytes (after Keilin and Hartree).

The source of these contradictions, as is now clear, was purely experimental difficulties connected with the study of the absorption spectrum of hemoglobin (and chlorophyll) in its natural state.

With the present state of experimental technique it is not at all difficult to measure with sufficient accuracy the visible absorption spectrum of a colored substance when it constitutes a transparent, homogeneous layer (a solution, a colored film, etc.). But a suspension of erythrocytes—in blood plasma or in physiological solution—constitutes an optically inhomogeneous and therefore strongly scattering (turbid) medium. Making measurements in such a medium is difficult not only because of the great losses of light due to scattering, but also because the usual standard methods applicable to transparent media are here altogether unsuitable, since they can lead to errors of a qualitative character.

In view of this, in order to obtain correct results in the case of turbid media, the experiment must be arranged so that, as far as possible, all the light scattered by the turbid medium reaches the receiving instrument. In work carried out jointly by the author’s laboratory and the physicochemical laboratory of the Institute of Hematology and Blood Transfusion of the Academy of Medical Sciences in Moscow1, this condition was realized in a very simple way: a cuvette containing a suspension of erythrocytes was placed close against the sensitive surface of a selenium photoelement. A sharp image of the exit slit of the spectral monochromator, in the form of a narrow strip 0.05–0.1 mm wide, was projected onto the front surface of the cuvette. Owing to the peculiar laws governing the scattering of light by large particles, such as erythrocytes (disc diameter of the order of \(5\ \mu\)), almost all the light scattered in passing through the turbid suspension falls upon the sensitive surface of the photoelement, and only a negligible part of it is scattered backward and is not recorded by the receiving instrument.

Fig. 14. Absorption spectrum of an erythrocyte suspension before and after hemolysis.

When this simple method was applied to measuring the absorption spectrum of an erythrocyte suspension, it turned out that in reality no “disappearance” of the \(\gamma\)-band exists. The entire visible spectrum of oxyhemoglobin is observed both in the case of a transparent solution and in the case of a suspension of erythrocytes.

For comparison of the spectral properties of hemoglobin in solution and inside the red blood corpuscle, it is important to be able to study the absorption spectrum in both cases while keeping all other conditions identical. Such a possibility is provided by the remarkable phenomenon of hemolysis, in which hemoglobin, under the influence of a number of external conditions, passes from the red blood corpuscle into the surrounding liquid. It is especially convenient to bring about this phenomenon by means of certain surface-active substances (namely, detergents); thus, for example, the addition of a negligible amount of saponin instantly “clarifies” blood or a suspension of erythrocytes, owing to the dissolution of hemoglobin in the blood plasma or in physiological solution (a \(0.9\%\) solution of NaCl in water), in

in which the erythrocytes are found. Making use of this phenomenon, it was possible to compare the absorption spectra of hemoglobin in solution and in the natural state in the following way: the absorption spectrum of a suspension of erythrocytes was measured by the method described above; then a grain of saponin was added to the cuvette, which caused hemolysis, and the absorption spectrum of the resulting transparent solution was measured with all the other conditions strictly maintained. The two curves obtained (Fig. 14) permit the following conclusion: 1) the maxima of the absorption bands of oxyhemoglobin in solution and inside the erythrocyte coincide; consequently, oxyhemoglobin, when extracted from the erythrocyte, undergoes no substantial changes; 2) however, the two spectra show a discrepancy exceeding the limits of experimental error. This discrepancy consists in the fact that in the weakly absorbed parts of the spectrum the curve rises, while in the strongly absorbed parts it falls. Obviously, here the very same effect which was taken for the “disappearance” of band $\gamma$ is manifested to a slight degree. Indeed, it is sufficient to move the cuvette a little away from the surface of the photoelement for the discrepancy between the two curves to increase sharply. When it is moved away to a sufficient distance, the entire spectrum is smoothed out altogether. This clearly shows that the causes of the peculiar features of the absorption spectrum of erythrocytes described above are rooted in the laws of light scattering in an optically inhomogeneous medium with colored elements. The study of light scattering in such media fully confirms this.$^{18}$ We shall not, however, dwell on this question here.

V

The question of the nature of chlorophyll in the natural state, i.e. in the chloroplast, is exceptionally important for elucidating the mechanism of photosynthesis. To orient oneself in this question, one should first of all turn to the study of the absorption spectrum of chlorophyll in extracts and in chloroplasts. Although methods for obtaining suspensions of undamaged chloroplasts have been described in the literature,$^{19,20,21}$ we know of no systematically accurate measurements of their absorption spectra. Usually the absorption spectrum of a plant leaf is measured, or of suspensions obtained by grinding leaves in water. From the experimental point of view, measurement of the absorption spectrum in an optically inhomogeneous medium is a considerably more difficult problem than measurement of transparent solutions. It is therefore not surprising that until very recently there were no more or less accurate data on the absorption spectra of the leaf in the natural state. In Fig. 15, as an example, the absorption spectrum of an ash leaf$^{22}$ is given and, for comparison, the absorption spectrum of an extract of its pigments (the blue band is not shown here). When the two spectra are compared, their difference is clearly visible. It consists above all in the fact that the absorption spectrum of the whole leaf ce-

shifted toward longer wavelengths, the magnitude of this shift for the principal red band being 120 Å. The cause of the shift was the subject of numerous discussions, in which the difference between the two spectra was associated either with a difference in the chemical nature of chlorophyll in extracts and in the chloroplast, or with its state in the chloroplast. Concerning the state of chlorophyll, the most varied hypotheses were advanced*). The hypothesis of the colloidal state of chlorophyll enjoyed the greatest popularity. This hypothesis was put forward by Ivanovsky and Gerlich and

Fig. 15. Absorption spectrum of an ash leaf and of an extract of its pigments.

was supported by Willstätter and Stoll. The basis for it was the similarity between the absorption spectra of leaves and colloidal solutions of chlorophyll, which also exhibit a shift toward the red.

On the other hand, Tswett, and subsequently Lubimenko, supposed that chlorophyll in its natural state, like hemoglobin, is a combination of a dye with a protein substance. Lubimenko, in particular, showed that when pigments are extracted from leaves of the plant Aspidistra elatior with water, there is obtained a pro-

*) A detailed review is given in the book by Willstätter and Stoll, “Untersuchungen über die Assimilation der Kohlensäure,” 1918.

2*

a transparent solution, the spectrum of which is “identical with the spectrum of the leaf.” Recent works in which, in parallel with spectral investigations, sedimentation under ultracentrifugation was studied apparently confirm the existence of a combination of chlorophyll with protein; however, they do not make it possible to decide whether this combination is a compound of constant stoichiometric composition (as is the case with hemoglobin) or whether here we are dealing with an adsorption bond. A more detailed consideration of the important questions touched upon here should be sought in the original works and in special monographs. (See, for example, the monograph by E. Rabinowitch[^11].)

VI

Let us consider another interesting example of the practical use of spectroscopic methods for investigations on a living organism. It turns out that certain features of the infrared absorption spectra of various hemoglobin derivatives open up great possibilities for quantitative experiment in the fields of physiology, pathology, and hygiene.

Already in 1935 Egert[^23] showed that hemoglobin derivatives behave in substantially different ways with respect to infrared (IR) rays. Especially sharply distinguished is carbonyl-hemoglobin HbCO, present in blood poisoned by carbon monoxide. Specifically for infrared rays, beginning at approximately \(800\,m\mu\) and farther toward longer wavelengths, HbCO is completely transparent, whereas Hb and HbO\(_2\) in this region possess appreciable absorption. This feature of HbCO was demonstrated by Egert in the following effective qualitative experiment. Into four cuvettes \(10\,\mathrm{mm}\) thick were poured: 1) reduced Hb (obtained by adding several drops of ammonium sulfide to \(20\,\mathrm{cm}^3\) of diluted hemolyzed blood); 2) HbCO (obtained by passing illuminating gas through); 3) HbO\(_2\); 4) “regenerated” Hb (obtained by passing air through HbCO). Behind the cuvettes was placed a white sheet of paper with an inscription, and the cuvettes were photographed. In Fig. 16 the first row is a photograph in visible light (orthochromatic plates). As can be seen, all the cuvettes here look equally opaque (black). The last row is a photograph in “invisible light,” i.e., on plates sensitized to the IR part of the spectrum, with illumination by an incandescent lamp through a black filter transmitting only IR rays. The complete transparency of HbCO for these rays is seen quite clearly.

Egert’s experiments were repeated and extended by Merkelbach[^24]. In Fig. 17 two pairs of Merkelbach’s photographs are shown. In the lower pair two drops of blood are photographed, one of which contained HbCO and the other HbO\(_2\). On the left is a photograph on panchro-

matic plate. Both drops on it appear equally black. On the right is shown a photograph taken under illumination by infrared rays (black filter) on infrachromatic plates. Here the drop containing HbCO is white, while the drop containing HbO₂ is still black. In the upper pair the same photographs were made with hemolyzed (i.e., transparent) blood. It is clearly seen that the drop containing HbCO is completely transparent.

Text in the figure:
Orthochromatic image
Region of long waves
Region of long waves
Region of long waves

1 — Defibrinated blood
2 — Blood containing CO
3 — Oxidized blood
4 — Reduced blood

Fig. 16. Photographs of cuvettes filled with solutions of blood treated in different ways, in different regions of the spectrum.

This property of HbCO makes it possible to carry out a very sensitive test for the presence of CO in the blood. Eggert reports that even after a patient, slightly poisoned by CO, had spent 15 minutes in fresh air and no traces of CO could be detected in the blood by the ordinary spectroscopic method (in the green part of the spectrum), an infrared photograph still gave a positive result.

A qualitative method of this kind, however, does not reveal many essential spectral properties of the various derivatives of Hb

Fig. 17. Oxygenated blood and blood poisoned with CO in visible and infrared light (according to Merkelbach).

in the i.-r. part of the spectrum. It was therefore important to pass from qualitative experiments to quantitative measurements of i.-r. absorption spectra. Such measurements had already been made approximately by Eggert himself. However, his results were very rough and rather schematic. Subsequently a number of investigators made systematic measurements. The most accurate results were obtained by Horecker \(^{25}\) with the aid of a photoelectric method. In Fig. 18, taken from Horecker’s work, three curves of the absorption spectra of Hb, HbO\(_2\), and HbCO are given. Here the following can be seen: a) the absorption of HbCO, beginning at 800 \(m\mu\) toward longer wavelengths, is equal to zero; b) HbO\(_2\) has a broad absorption band with a maximum at 920 \(m\mu\); c) the absorption curves

Fig. 18. Infrared absorption spectra of hemoglobin, oxyhemoglobin, and carboxyhemoglobin (according to Horecker).

...absorption of Hb and HbO₂ at 800 mμ intersect, which shows that for this wavelength both substances have the same absorption; d) in the region near 700 mμ, i.e. in the far red part of the spectrum, the difference between the absorption of Hb and the absorption of HbO₂ is especially great. Knowing the absorption spectra of various derivatives of Hb, it is possible to analyze blood for the relative content of reduced Hb, HbO₂, and other derivatives. However, owing to the easy oxidizability of Hb, carrying out the various necessary operations in a cuvette without disturbing the composition of the blood is associated with great difficulties. Moreover, it is of particular interest to be able to make observations on the composition of the blood continuously and directly in the living organism. In view of this, a number of works appeared in which investigators were engaged in developing a method for the analysis of blood in vivo. The most convenient method for determining the relative concentration of Hb and HbO₂ consists in measuring the absorption of light of a definite wavelength that has passed through tissue of the human body penetrated by blood vessels. Usually the ear is used for this purpose—its lobe or the upper part of the auricle. As the spectral region, the region near 700 mμ is usually chosen, where the differences between the absorption of Hb and HbO₂ are especially great. To isolate this region, light filters were used, i.e. appropriately colored glasses or films, and the intensity of the transmitted light was measured by a photoelement—usually a selenium one.

Here, however, the following difficulty arose: the absorption of light changes not only as a result of a change in the ratio \(\frac{\mathrm{HbO}_2}{\mathrm{Hb}}\), but also as a result of fluctuations in the blood filling of the vessels. At first they tried to take into account the influence of these fluctuations by observing the blood pressure with the aid of an ordinary manometric method. But this proved extremely complicated. In view of this, Mattes and Gross\(^{26}\), and subsequently, in a still more ingenious and elegant form, Millikan\(^{27}\), used the principle of “two-color colorimetry.” The idea is as follows: since at 800 mμ the absorption curves of Hb and HbO₂ intersect, absorption in this region characterizes the total amount of Hb (together with HbO₂) filling the vessels in the given region of tissue. If, therefore, with the aid of two light filters one isolates approximately 700 mμ and, in addition, separately the so-called k-region, then the change in transmission of the latter will characterize fluctuations of blood pressure, while the change in transmission in the 700 mμ region will make it possible to determine the ratio \(\frac{\mathrm{Hb}}{\mathrm{HbO}_2}\). Mattes and Gross used two photoelements selected according to their characteristics and provided with the filters indicated above; Millikan used one differential photoelement and an arrangement in which fluctuations of blood pressure were automatically eliminated. In Fig. 19 a schematic example is given...

recording of one of the experiments of Mattes and Gross. Here the lower curve gives the recording of blood pressure, measured manometrically; the next curve gives the “i.-k. photocurrent,” i.e., precisely that photocurrent whose magnitude depends on the presence of “total hemoglobin” and which therefore also represents a blood-pressure curve, recorded, however, by a photoelectric method. The third curve from below gives the “red photocurrent,” i.e., the photocurrent characterizing simultaneously the total amount of Hb and the ratio of HbO$_2$ to Hb. Finally, the upper curve gives respiration. Let us pay attention to the “red” photocurrent curve. This curve shows a noticeable decline beginning from the moment when the subject began to breathe air with a reduced oxygen content (7% O$_2$).

Fig. 19. Spectroscopic recording of changes in blood composition (after Mattes and Gross).

With reduced oxygen content (7% O$_2$). Since, however, the “i.-k. photocurrent” curve does not change, the decline of the “red” curve should be attributed entirely to a decrease in the concentration of HbO$_2$ in the blood and to the associated increase in absorption for $\lambda \sim 700\,m\mu$. In the region $P$ between arrows 2 and 3 the subject held his breath for a short interval of time. This was reflected in the “red” curve by a sharp rise. However, the rise is also exhibited by the “i.-k.” photocurrent, which indicates a simultaneous change in the blood filling of the vessels, confirmed by the blood-pressure curve. Thus, in the present case, one cannot judge the ratio Hb/HbO$_2$ from the change of the “red” curve.

Millikan, as already indicated, used, for analogous measurements, a single selenium photocell, which was divided into two parts and acted as two photocells connected in opposition to each other. By placing in front of these two halves two appropriately selected light filters, he automatically eliminated the change in photocurrent due to fluctuations in blood filling. On this principle Millikan built the “oximeter” instrument, with the aid of which, with extraordinary simplicity, it is possible continuously to monitor the magnitude of the ratio Hb/HbO$_2$, which is very important in many

cases of physiological, medical, and other investigations. The instrument weighs only 30 g and is fastened to the upper part of the auricle.

In all the investigations discussed so far, it was tacitly assumed that the tissue of the human body itself contributes nothing to the absorption of light, apart, of course, from strong scattering. Indeed, it has long been known qualitatively that tissues possess a high degree of transparency for the red and infrared regions of the spectrum. However, the quantitative data available in the literature on the absorption spectrum of living tissue taken as a whole are contradictory:

Fig. 20. Infrared transmission spectra of tissues of the human body28.
1—earlobe; 2 and 3—cheek.

for some authors (Bachem and Reed) the maximum transmission occurs at 800 mµ, for others (Cartwright) at 1100 mµ. The absolute values of transmission were given for certain definite wavelengths selected by filters, and are likewise contradictory.

A systematic study of the absorption spectrum of tissues of the human body (ear and cheek) in the extreme red and near-infrared region of the spectrum was recently carried out in the author’s laboratory by A. A. Il’ina28. In doing so, possible sources of error were taken into account (light scattering, heating of the photocell, etc.), and the curves were measured in detail at intervals of every 5–10 mµ. Examples of such curves are shown in Fig. 20. As can be seen, maxima (at 700–800 mµ) and minima (at 980 mµ) of transmission appear on these curves with great distinctness. The interpretation of these maxima and minima also presents no difficulty. The sharp minimum at 980 mµ and the steep fall of transmission at \(\lambda > 1200\) mµ are undoubtedly due to the presence of water in the tissues. In the region 600–800 mµ, although

curve is determined by the spectral properties of blood. The noticeable scatter of points in the region \(640—760\,m\mu\), obtained on all the curves, is not the result of random errors, but is due to fluctuations in the value of the ratio \(\mathrm{HbO_2}/\mathrm{Hb}\), which must have an especially strong effect precisely in this region. The entire curve could be reproduced rather accurately artificially, by placing, in front of the photoelement, cuvettes with blood and water at definite layer thicknesses (Fig. 21).

This makes it possible to think that, along with spectral methods for determining the degree of oxidation of blood and the spectral method for determining

Fig. 21. Artificial reproduction of the infrared absorption spectrum of tissues[^28].

blood pressure, it is possible to develop a spectral method for determining the water content in tissues, which is of considerable interest for physiology and pathology.

VII

In this brief survey there is no possibility of dwelling on other—very diverse—applications of spectroscopy in biology. We shall confine ourselves, in addition to what has been said, to only a few remarks. Absorption spectroscopy is at present the most important auxiliary means in the identification and study of the structure of synthetically obtained hormones (see Morton’s monograph on this subject). There are also some, although not very extensive, results concerning the spectroscopy of protein molecules. As is known, the structural units of these giant molecules are amino acids; the number of different amino acids occurring in proteins is not less than 25. Their qualitative and quantitative determination is a very important problem, so far solved only for a few amino acids. In particular, the cyclic amino acids—tryptophan and tyrosine—give bands of spectral absorp-

absorption in the ultraviolet in the region of 280 mμ. It is significant that the absorption of a protein is approximately additively composed of the absorptions of the individual amino acids. Attempts at spectroscopic analysis of proteins are based on this, giving results close to the data of chemical analysis, although not fully coinciding with them. Obviously, possibilities for analysis exist here, and further work is required in order to improve the methods (for details see Morton’s monograph, ch. IX).

We shall give a few more examples, which we borrow from Twyman and Allsopp (ref. A IX).

“Using the extinction coefficient of pseudoglobulin, Marrack and Campbell-Smith investigated the nature of the floccules of diphtheria toxin and antitoxin (the so-called Ramon precipitate). The results apparently show that such an immunity reaction has the nature of a specific chemical reaction, and indicate a new approach to problems of this type. It should be noted that in this method one has to deal with only 2–3 mg of protein, the concentration of which can be determined with an accuracy of 3–4%.

Another application is the method of determining purines. Their extinction coefficient is very high, so that they can be determined in extremely strongly diluted solutions: a solution of uric acid containing 2.6 mg in 100 cm³ can be determined with an error of about 3%, and approximately one tenth of this concentration can be detected in solution—in all other respects optically transparent.”

This entire interesting and important field has only in recent years begun to attract attention from biologists and chemists, as well as from physicists*). It is an extensive field for research, as yet little developed. Significant successes should be expected in connection with the development of experimental technique. In this respect, of particular importance is the replacement of laborious and insufficiently accurate methods of photographic photometry in the ultraviolet by methods of photoelectric spectrophotometry (a survey of these methods will be given in another article), which combine a high degree of accuracy with rapid operation. The appearance of technically perfected instruments, whose operation is as simple as the operation of amateur radio receivers, makes it possible to hope that,

) A sign of the growing attention to biological problems on the part of physicists may be the appearance of a number of articles in purely physical journals. Thus, in the journal Reviews of Modern Physics, in addition to the review by Lefebvre mentioned above (see bibliography A V), two articles have recently appeared: by Hewson, “On the Biological Action of Radiation,” and by J. Frank, “On Photosynthesis by Means of Isolated Chloroplasts” (Rev. Modern Physics, 17, Nos. 2–3, 1945). In the same journal a review by J. Frank on photosynthesis has been promised. A number of works appear from time to time in the Journal of Applied Physics*.

that ultraviolet spectrophotometers will attain in chemical, biological, and medical laboratories the same wide dissemination as the Pulfrich visual photometer has already attained.

In another direction, significant successes in solving the most subtle problems of biology should also be expected in connection with the entirely new development that ultraviolet microscopy has received (ultimately based on differences in the ultraviolet absorption spectra of organic substances), thanks to the work of E. M. Brumberg (State Optical Institute in Leningrad). The successful solution of the problem of combining, by means of a single instrument, microscopic and spectral investigations in the ultraviolet, recently accomplished by E. M. Brumberg, very greatly increases the power of this remarkable method.

LITERATURE

A. Monographs

The following monographs and surveys provide exhaustive references to the original literature:

I. R. Morton, The Applications of Absorption Spectra to the Study of Vitamins, Hormones and Coenzymes, Adam Hilger Ltd., London, 1942.
II. E. S. Miller, Quantitative Biological Spectroscopy, 2-nd Edition, Burgess Publishing Co., Minneapolis Minn., 1940.
III. W. R. Brode, Chemical Spectroscopy, 2-nd Edition, New York, 1945.
IV. H. Ley, Beziehungen zwischen Absorption und chemischer Konstitution. Handbuch der Physik, Vol. XXI, p. 57.
V. J. Loutbourrow, Borderland Problems in Biology and Physics. Reviews of Modern Physics, 12, 267, 1940 (contains a detailed index of the original literature).
VI. R. B. Barnes, R. G. Gore, U. Liddel and Van Zandt Williams, Infrared Spectroscopy, New York, 1944.
VII. Gladys A. Anslow, Ultraviolet Spectra of Biological Important Molecules. J. Appl. Physics, 16, 41, 1945.
VIII. Heilmeyer, Spectrophotometry in Medicine, A. Hilger Ltd, 1944.
IX. Twyman and Allsop, The Practice of Absorption Spectrophotometry, A. Hilger Ltd.

B. Literature cited in the text

1. G. Scheibe and W. Frömel, Hand- und Jahrbuch der Chem. Physik, 1936, B. 9, Abschn. III—IV, p. 157.
2. V. A. Anri, Study of the Absorption of Ultraviolet and Infrared Rays in Connection with the Structure of Molecules, Leningrad, 1919.
3. H. Sponer and E. Teller, Rev. Modern Physics, 13, 76, 1941. There, too, are further examples.
4. K. W. Hausser, R. Kuhn, A. Smakula u. M. Hoffer, Z. physikal. Ch., B. 29, 371, 1935; K. W. Hausser, R. Kuhn, A. Smakula, K. Kreuchen, Z. physikal. Ch., B. 29, 363, 1935; K. W. Hausser, R. Kuhn, A. Smakula, W. Deutsch, Z. physikal. Ch., B. 29, 378, 1935; K. W. Hausser, R. Kuhn, A. Smakula, Z. physikal. Ch., B. 29, 384, 1935.
5. R. Mulliken, Rev. Modern Physics, 14, 265, 1942 orka
6. R. Pohl, Naturwiss., 15, 435, 1927.

1. E. Rabinowitch, Photosynthesis, New York, 1945.
2. A. Stern and H. Wenderlein, Z. physikal. Ch., A 174, 81, 1935.
3. A. Stern and H. Wenderlein and H. Molvig, Z. physikal. Ch., A 177, 40, 1936.
4. A. Stern and H. Wenderlein, Z. physikal. Ch., A 176, 81, 1936.
5. E. Rabinowitch, Rev. Modern Physics, 16, 226, 1944.
6. F. P. Zscheile Jr., Cold Spring Harbor Symposia on Quantitative Biology, vol. III, p. 108, 1935.
7. A. B. Macallum and R. C. Bradley, Science, 71, 341, 1930.
8. G. A. Adams, R. C. Bradley and A. B. Macallum, Bioch. m. Journ., 28, 482, 1934.
9. G. A. Adams, Biochem. Journ., 32, 646, 1938.
10. D. Keilin and E. F. Hartree, Nature, July 19, No. 374, 75, 1940.
11. A. A. Il'ina, Kh. M. Ravikovich, D. L. Rubinshtein, E. V. Shpol'skii, DAN, 48, 346, 1945.
12. E. V. Shpol'skii and A. A. Il'ina, Acta Physicochimica URSS, 1946 (in press).
13. R. Hill, Proc. Roy. Soc., London, B, 127, 192, 1939.
14. R. Hill and R. Scarisbrick, Proc. Roy. Soc., London, B, 129, 238, 1940.
15. J. Franck, Rev. Modern Physics, 17, 112, 1945.
16. A. Il'ina, Journal of Physical Chemistry, 1946 (in press).
17. J. Eggert, Naturwiss., 23, 281, 1935.
18. Merklebach, Schweiz. Med. Wochenschr., 65, 1142, 1935.
19. B. Horecker, J. Biolog. Chem., 148, 173, 1943.
20. Mattes and Gross, Arch. exp. Pathol. u. Pharmakol., 191, 523, 1939.
21. G. A. Millican, Rev. Sci. Instr., 13, 434, 1942.
22. A. Il'ina, DAN, 1946 (in press).
23. E. V. Shpol'skii, Spectroscopy of chlorophyll. Proceedings of the All-Union Conference on Photosynthesis, 1946 (in press).

1. Reference number as printed on the page. ↩