Download - INFLUENCE MEASUREMENTS · cellular components in suspension scatter light with astrong spectral selectivity (5, 6). Scattering maxi-maoccur on the long wave length sides at absorption

INFLUENCE OF SELECTIVE LIGHT SCATTERING ON MEASUREMENTS OFABSORPTION SPECTRA OF CHLORELLA 1, 2

PAUL LATIMERDEPARTAIENT OF PLANT BIOLOGY, CARNEGIE INSTITUTION OF WASHINGTON STANFORD, CALIFORNIA; AND

DEPARTMENT OF PHYSICS AND HOWARD HUGHES MEDICAL INSTITUTE, VANDERBILT UNIVERSITY,NASHVILLE, TENNESSEE

Experimental information about the light scatter-ing characteristics of biological cells is surprisinglyscarce. Even though all measurements of absorp-tion spectra of intact biological structures must beinfluenced to some extent by scattering, there havebeen few investigations dealing specifically with scat-tering by cell suspensions. However, recent experi-ments have revealed that light absorbing cells andcellular components in suspension scatter light witha strong spectral selectivity (5, 6). Scattering maxi-ma occur on the long wave length sides at absorptionmaxima. Other studies have demonstrated that thisselective scattering can influence measurements ofabsorption spectra, shifting apparent band positionsby 10 to 15 m,u (7, 8). The present paper describesfurther investigations of phenomena related to selectivescattering. We attempted to obtain information aboutlight scattered at small angles to an incident beamfrom suspensions of Chlorella pyrenoidosa Chick(Emerson strain) by using special experimentalmethods to nmeasure absorption spectra.

We have made use of the fact that the effects ofscattering on measurements of absorption spectra ofturbid cell suspensions depend on some of the experi-mental conditions. Specifically, we measured twoseries of absorption curves in which a single factor thatinfluences the effects of scattering was varied through-out a series. In the first case, absorption curves of acell suspension were measured with a special spectro-photometer which allowed the detector to view onlythat light coming from the cell suspension at certainpredetermined angles to the incident beam. In thesecond series of measurements, we fixed the geometryof the spectrophotometer so that the light detector re-ceived only directly transmitted light (parallel to theincident beam) and varied the index of refraction ofthe suspension medium from curve to curve (seeBarer (1)).

APPARATUS AND PROCEDUREIn order to measure absorption curves by allowing

the light detector to observe only light coming fromthe suspension at specified angles, we constructed aspecial spectrophotometer by modifying the fluores-

Received March 23, 19592 Parts of this investigation were assisted by a research

grant, (RG-5516) from the National Institutes of Health,Public Health Service.

cence spectrophotometer of French (4). The sampleholder and mirror system was replaced by a tungstenlight source, and baffles and a suspension vessel wereinserted between the exit slit and the light detector(Photomultiplier tube, R.C.A. 6217) (see fig 1).The baffles isolated a parallel (± 10) beam of lightfrom the monochromator output. A thin (1/8 inch)layer of suspension was used to reduce uncertaintiesin the angle of observation. A special baffle (angleselecting assembly) was constructed which alternatelypermitted the photomultiplier tube to observe 00 (di-rectly transmitted) light (through the small centeropening) or light scattered at an angle to the incidentbeam (through the doughnut shaped opening, thecenter of which was in line with the incident beam).The angle of observation could be increased by mov-ing the angle selecting assembly (and the photo-multiplier tube and diffusing plate) closer to the sus-pension vessel. The doughnut shaped opening wasused instead of a simple small hole for defining theangle of observation at non-zero angles since it allowedmore light to pass with the same angular resolution.The diffusing plate (made of opal glass) eliminatederrors which might be produced by differences in thesensitivity of different parts of the photocathode.

As a control, we also measured absorption curvesof the suspensions by placing diffusing plates at thesuspension and blank vessels as suggested by Shibata(11 ). Since the effects of scattering on this lattercurve were the least, and since this method has beenfound to yield absorption curves in general agreementwith those obtained by the more reliable integratingsphere technique, which further reduces the effects ofscattering (compare curves in references (5) and(7) ), we assume that absorption curves measured inthis way-are good approximations of the actual ab-sorption spectrum of the cell suspensions. For thesemeasurements, the angle selecting assemblv was re-moved, the diffusing plate was placed against thevessels, and the photomultiplier tube was moved to theleft.

To determine the spectral absorption curves, wemeasured separately (using the automatic sweep mech-anism) spectral curves of the light passing throughthe suspension (at a specified angle) and throughthe blank vessel (at 0°). Actually 1 curve was: (out-put of monochromator) X (sensitivity of light de-tector system) X (transmission of suspension) =f(), while the other was: (output ....) X (sensi-

193

www.plantphysiol.orgon April 14, 2020 - Published by Downloaded from Copyright © 1959 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

PLANT PHYSIOLOGY

tivity . . . . ) X (transmission of blank) = f'(X).Since the first 2 terms in the products are the same(see below), "absorbance" curves were obtained bydividing the levels of the curves at 5 mpt intervals andchanging to logarithmic coordinates (absorbance =log I./I).

The 2 types of curves were measured alternately,3 sweeps of the spectral region investigated beingmade in each case. The averages of the 3 curves ofeach type were used to calculate the curves shownin figure 3.

This method requires that the product, (output ofmonochromator) X (sensitivity of detector system),remain constant during the entire determination of 1curve. This condition was usually fulfilled as evi-denced by agreement between different curves of thesame quantity taken at different times. When suchcurves disagreed, the data were discarded.

Absorption curves of cells in suspension media ofdifferent indices of refraction were measured with aBeckman DK-2 recording spectrophotometer. Thenormal arrangement of its optical system was used tomeasure directly transmitted light (0±+30) (arrange-ment (a) in fig 2), while diffusing plates wereplaced after the suspension and blank vessels (against

MoNochromatorslit

I

the vessel holder) for the 00 to 90° curves (arrange-ment (c) in fig 2). Further details of this methodare given in references (7) and especially (8). Mediaof different indices of refraction were prepared by dis-solving bovine serum albumen (Armour's fraction V)or glycerol in water (buffer). Cells from the sameculture were added to each medium to yield suspensionsof equal cell density.

EXPERIMENTAL RESULTSDEPENDENCE OF ABSORPTION SPECTRA ON GEOME-

TRY OF OPTICAL SYSTEM: Experimental absorptioncurves in which the angle of observation of the lightdetector was varied are shown in figure 3. The uppercurves are absorption spectra measured with the photo-multiplier tube observing only light coming from thesuspension at 0°, 3.4°, and 320 respectively. Thecurves are quite different, not only in shape but alsoin the position of the "absorption" maximum (688 m,at 00, 682 mu at 3.40 and 668 m,u at 320). Othermeasurements show this shift to be a gradual one withangle. Although we were not able to extend themeasurements to larger angles with this apparatus,the results of our previous measurements of lightscattered at 900 are very nearly comparable (5, 6).

PhotomrltipliertubeBI.ank

BafflesAngle selecting assembly

/6,\\\

1" 1"l

4 ,j8j 2-FIG. 1. Apparatus for measuring spectral curves of light transmitted or scattered at small angles to the incident

beam. The angle is varied by changing the distance between the suspension and the angle selecting assembly.

194

1 3 &

a

S 3 -

a3 "to g

"

8 a



LATIMER-SELECTIVE LIGHT SCATTERING BY CHLORELLA

Semple (S)

. . S*. _lo - S.:

Diffos1"g plot*

FIG. 2. Schematic diagramssystem of a spectrophotometer.

0.2

0.1

0.0

-

-0. II I s

600 650 700 7500.3

67241

0.1 _

*0.0 I I IW50

Wavltength ,1.

Photocell (P)

-E (a)

(b)

--EJ

of part of the optical

FIG. 3. Chlorella: upper curves, absorption spectradetermined by measuring light transmitted or scattered bya cell suspension at indicated angles with the apparatusin figure 1. The curves were arbitrarily normalized inheight. The indicated angles, 00, 3.40, and 320, are notcorrected for refraction at the vessel surface. This cor-

rection changes the angles to 00, 2.60, and 240 respectively.The iower curve was determined with the diffusing plateagainst the vessel, as in Figure 2 (c).

If the curve for 900 scattering by Chlorella were plot-ted as an absorptio curve (this would require invert-ing the curve and c}ing from linear to log units),the "absorption" ma in would appear at about 655m,u which is still further to the left.

The lower curve in figure 3 (maximum at about672 m,u) was obtained by placing the diffusing plateat the suspension vessel. It essentially represents a

weighted average of spectral curves (such as thoseshown above it) for light coming out of the suspensionat all angles of from 0° to 900 to the incident beamand with absorption maxima varying in position from688 m,u to about 655 mu.

DEPENDENCE OF ABSORPTION SPECTRUM ON INDEXOF REFRACTION OF SUSPENSION MEDIUM: Absorp-tion spectra of Chlorella in media of different indicesof refraction as measured with the Beckman DK-2spectrophotometer (using arrangements (a) and (c)in fig 2) are shown in figures 3 and 4. First, itshould be noted that the positions of the absorptionbands of all suspensions (except that for medium,n = 1.40) are different in the upper and lower seriesof curves (see reference (7) for a preliminary reportof this effect). For instance, the red absorption maxi-mum of cells in water (n = 1.33) is at 685 m,u in the(a) series of curves while it occurs at 675 mu in the(c) series.

As the index of refraction of the medium is in-creased towards approximately that of the cells, as-

sumed here to be 1.40, not only does the general heightof the curves decrease, but the apparent positions ofthe bands on the upper curves shifts towards the actualposition of the bands. On further increasing the in-dex of refraction of the medium to 1.42 (approximate-ly 40 % albumen), the shift to the left of the bandpositions continues so that the red band now appears

at 668 mu.

It should be noted that the positions of the bandson the lower curves in figure 4, which presumably rep-

resent the best approximation of the actual absorp-tion spectra, are not significantly altered by the medi-um. Therefore, the variations in the positions of thebands in the upper part of the figure must be attributedto optical phenomena and not to actual differences inthe true absorption spectra of the cells.

On the other hand, a similar series of measurementson Chlorella cells in glycerol-water solutions gavequite different results (see fig 5). While the generallevel of the upper curves is reduced somewhat as theindex of refraction of the medium increases, onlv a

slight shift in the apparent positions of the bands isobserved. Other curves in the same series for inter-mediate values of n were similar to those shown in thefigure. The general difference between the behaviorof the curves in figures 4 and 5 can be attributed todifferences in the nature of the additives, bovine serum

albumen and glycerol, respectively. The albumen,being a large molecule, does not penetrate the cellmembrane, and the osmotic pressure of its solutions is

195

J00



196 PLANT PHYSIOLOGY

Wuvile.vb,Wmp.4,, SOO 600

Wav..lugth'N'mJOO

FIG. 4 (left). Chlorella: absorption spectra of suspensions of equal cell density in different water solutions ofbovine serum albumen having the indicated refractive indices. The curves were measured with a Beckman DK-2recording spectrophotometer using arrangements (a) and (c) in figure 2.

FIG. 5 (right). Chlorella: absorption spectra of suspensions of equal cell density of water-glycerol solutionshaving the indicated refractive indices.

probably not very different from that of water. How-ever, the small glycerol molecule readily penetratesthe cell membrane and the osmotic pressure of itssolutions is high. Hence, the optical properties of thecells in glycerol-water solutions must be different fromthose of cells in water or albumen solutions becauseof penetration of the glycerol and the exit of waterfrom the cell. Both processes increase the index ofrefraction of the cells. Scattering depends on the dif-ference between the index of refraction of the cellsand that of the surrounding medium (see next sec-tion). It is suggested that the addition of glycerolto the medium did not effect large changes in this dif-ference since both indices are raised by it.

As noted elsewhere (8), the positions of the maxi-ma on curves measured with arrangement (a), figure2, vary slightly with suspension concentration (pre-sumably because some of the observed light has en-countered 2 or more cells). For cells in water, thethinnest suspensions gave peaks displaced furtherest tothe right. We estimate that on extrapolating to zeroconcentration, the band maximum at 685 m,u in theupper part of figure 4 would be shifted to about 689 m,.

THEORETICAL INTERPRETATION

Many similarities exist between the angular de-pendent curves in figure 3 and the medium dependentcurves in figure 4. In both cases, the position of the"absorption" maximum is found on either side -of theactual maximum depending on experimental condi-tions. Both sets of results are clearly related to thephenomena of selective scattering which can be ex-plained in terms of the theory of scattering by absorb-ing particles (9).

To a first approximation, total scattering by bothabsorbing and non-absorbing particles depends on theindex of refraction (relative to the surrounding medi-um) of the scattering particle according to the follow-ing equation:

S = k (m-1)2 (1)where S is scattering per unit incident light, k is aconstant and m is the relative index of refraction ofthe particle. (If the index of refraction of cells is 1.40and that of the medium (water) is 1.33, the relativeindex of refraction of cells is 1.40/1.33 = 1.05.)According to the theory of anomalous dispersion, m



LATIMER-SELECTIVE LIGHT SCATTERING BY CHLORELLA

becomes a complex number in regions of absorption(m = n-in' where we call n the real part and n' theimaginary part). Both parts vary rapidly and in acharacteristic manner with wave length. The imagi-nary part of m is very nearly proportional to absorp-tion while the real part varies assymetrically aboutthe absorption band. (See figure 6 A for theoreticalcurves of these 2 parts for Chlorella.) Since the realpart of m passes through a maximum on the longwave side of the absorption band, S (in equation 1)should also pass through a maximum on the long wavelength side if m > 1. This effect which we calledselective scattering, was observed in studies of 90°scattering of several different biological systems.

It would be desirable in this section to use an exacttheory of scattering by colored particles to calculatetheoretical curves for comparison with all of our ex-perimental curves. While it appears that the Mietheory of scattering by light absorbing particles shouldbe capable of explaining all of our results, the calcula-tions for absorbing particles of this size would be al-most prohibitively complex. On the basis of limitedefforts to apply the theory to Chlorella, we estimatethat the determination of a single point on one spec-tral curve would involve several thousand calculations.

On the other hand, van de Hulst (12) has obtainedapproximate expressions from the rigorous MIie theory

and the classical theory of dispersion which predictextinction and absorption by some homogeneousspheres. The equations, as restated in reference (9),are useful in predicting extinction and total scatteringby nearly spherical biological cells. We used it here,in a further modified form which eliminates some ofthe approximations, to calculate extinction curvesfor Chlorella cells in media of different indices ofrefraction. Since the upper curves of figure 4, forChlorella in bovine serum albumen solutions, areactually extinction curves (extinction = absorption+ total scattering), the theoretical curves should agreereasonably well with these experimental curves if,indeed the theory is applicable to suspensions ofChlorella cells.A Chlorella cell was assumed to be a homogeneous

sphere, diameter: 3.2 ,. We assumed its average in-dex of refraction at wave lengths remote from absorp-tion bands to be 1.40. The index in regions of absorp-tion was calculated from the classical theory of dis-persion using the extinction coefficient of cells (seereference (9) ) and a band half-width of the red ab-sorption band of 33 m,u. The absorption curve wasassumed to be symmetrical about 675 m,u. No allow-ance was made for the shoulder at 650 mti which ispresumably caused by chlorophyll b.

The theoretical extinction curves are shown in

*640 680 720 600 640 680WAVELENGTH, m/,,, WAVELENGTH, m}i

FIG. 6. Chlorella: theoretical curves. (A) Real (n) and imaginary (n') parts of average index of refractionof entire cell (2). (B) Extinction of cells in media of indicated refractive indices, calculated using (A) above and(12).

197



PLANT PHYSIOLOGY

figure 6 B. Comparison with the corresponding ex-

perimental curves in the upper part of figure 4 re-

veals that the observed dependence of the band posi-tions on the index of refraction of the medium is clear-ly predicted by theory. The bands on the theoreticalcurves at 689, 684, 675, and 669 mu are in good agree-

ment with those on the corresponding experimentalcurves at 685, 683, 677, and 668 mu respectively. Itis likely that the discrepancy between band positions,689 and 685 m,u for cells in medium, n = 1.33, is

caused by effects of multiple scattering on the experi-mental curve as previously mentioned. The differencebetween the band positions of 677 and 675 mu for cellsin a medium of n = 1.40 may be caused by our assum-

ing for the theoretical calculations, a value (1.40) ofthe index of refraction of the cells (at wave lengthsremote from the absorption band) which is too low.

On the other hand, the general levels of the experi-mental curves vary more with the index of refractionof the medium than do those of the theoretical curves.

This difference between experiment and theory may

be caused by the following factors. The logarithmicordinate scale was used for the experimental curves

while the scale for the theoretical curves is linear.Furthermore, thus far we have made no allowancefor effects of scattering by non-absorbing componentsof the cells. While it is difficult to say how the omis-sion of this factor would influence the results, it mighteasily produce the above mentioned discrepancy be-tween experiment and theory.

There is also another less rigorous, but more

tangible, theoretical explanation of the upper curves

in figure 4. It might be suggested that the displace-ment of the absorption band of Chlorella in water(when measured at 00) from its actual position at675 to 685 mnt can be explained in terms of thefact that we actually measured extinction whichis the sum of absorption plus total scattering.Thus if the actual absorption band lies at 675 mu andthe scattering band at 690 mnA (see reference (5)), a

suitable additive combination of the 2 bands wouldyield an extinction band at 685 m,ix. As one increasesthe index of refraction of the medium (to 1.36 and1.40) toward that of the cell (about 1.40), the magni-tude of scattering (see equation 1) and its contribu-tion to the extinction band is reduced, and the band isshifted from 685 to 683 my/ to about 675 mFL. Now,passing on to a medium (n = 1.42) of index of refrac-tion greater than that of the cells, we see that (m-1)in equation 1 becomes a negative number (although(m-1)2 will nevertheless be positive). In this case,

the index of refraction of the cells (see fig 6 A)has a minimilunm on the short wave length side of theabsorption band, so that the quantity (m-1) has a

minimum value since m < 1 but nevertheless (mi- 1)2has a maximum value. In this case there should bea scattering maximum on the short wave length sideof the band instead of the long wave length side. Thusthe extinction would be the sum of the absorption(maximum at 675 mMi) and scattering (peak betweenabout 659 and 675 my). And indleed, the maximum

extinction does seem to represent a sum of 2 suchcomponents with a maximum at 668 m,u.

On the other hand, if we consider the upper curvesin figure 3 for the angular dependence of the absorp-tion, neither equation 1, van de Hulst's theory, or theabove considerations are applicable, and no theoreticalexplanation is offered at this time. However, someunderstanding of the observed phenomena of angulardependence of band position may be obtained by con-sidering that advanced by Lothian and Lewis (10) toexplain the results of their spectral curves of red bloodcells. These authors reported interesting extinctionand absorption curves for red blood cells which weremeasured by methods similar to both of ours forChlorella in water, n = 1.33 in figure 3. (Theirspectral curves have been substantially reproducedunder special conditions in our laboratory by J. A.McClure.) While their data might in part be in-terpreted in terms of a peak shift caused by selectivescattering as we suggested above, they explained thedifferences in their curves in terms of interferenceand diffraction effects. The explanation involves in-terference between light passing through the cellwith that passing around it.

We have previously described absorption spectrameasured with a diffusing plate at the suspensionvessel both as reasonable approximations of the actualabsorption spectra and also as a composite of anumber of different spectral curves such as those infigure 3. It doesn't seem immediately obvious howthe diffusing plate curves can be these two things atonce. However, the explanation appears to lie in theinterpretation of the curves in figure 3 in terms of in-terference and diffraction effects.

DISCUSSION AND CONCLUSIONS

By selecting the experimental conditions for meas-uring absorption spectra of the green alga Chlorellapyrenoidosa, we found that the positions of the absorp-tion bands vary with the conditions over rather widewave length intervals. The positions of band maxi-ma as well as the shapes of the absorption curves arefunctions of both the index of refraction of the sur-rounding medium and the geometry of the opticalsystem of the spectrophotometer. These effects arestrictly optical in nature and do not reflect actualchanges in the absorption spectra of the cells, whichaccording to curves measured with diffusing plates,remained constant through all measurements.

Actually it should be noted that the position of thered absorption band (672 to 675 my) as measuredwith diffusing plates in figures 3 and 4 is slightly dif-ferent in the 2 cases. The cells used in these 2 meas-urements were from different cultures. We believethat this difference is a real one although it might beaccounted for in other ways. For instance, differentspectrophotometers were used for the 2 sets of meas-urements and there may have been slight calibrationalerrors. Also, the diffusing plate does not collect allscattered light from the suspension but only that

198



LATIMER-SELECTIVE LIGHT SCATTERING BY CHLORELLAemerging in the forward direction. The level ofcurves at 740 mjvwhere little if any absorption occursis probably caused by loss of light caused by back scat-tering. It is possible that differences in the scatteringcharacteristics of the cells from different cultures soeffected the back scattered light that the apparent bandposition was shifted.

It may well be that many of our curves measuredin special ways (upper parts of figs 3, 4, and 5)should not actually be called absorption spectra. Thisillustrates the fact that while absorption spectra ofnon-scattering dyes in solution have a rigorous op-erational definition, no counterpart exists for turbidsuspensions. However, in normal measurements ofabsorption spectra of biological cells and structures,we attempt to obtain curves which approximate thosewhich would be obtained if the light absorbing com-pounds, in the same physical and chemical state, wereuniformly distributed in a medium such that no scat-tering occurred. Unfortunately, these experimentalconditions cannot be attained even in an integratingsphere which detects a representative sample of alllight scattered out of the suspension vessel. While theintegrating sphere appears to provide the most reliablemeasurements of absorption spectra, results obtainedwith it, as well as by most other methods, are never-theless influenced by second order effects of scatteringwithin the suspension (such as alteration of the ef-fective path length of light through the suspension).Another factor which distorts absorption curves is themutual shading effect which has been described theo-retically for non-scattering particulate suspensions byDuysens (3) and E. E. Jacobs (the effect flattensabsorption bands).

The optical effects reportedI here, and also thosementioned above, point to the fact that the eventswhich occur when a beam of light enters a suspensionof biological cells are still poorly understood. It ap-pears at the present that insteadl of attempting to de-vise, by empirical means, a better experimental methodof measuring absorption spectra of cell suspensions,it is more appropriate to first gain a clearer under-standing of the events which occur on the passage oflight through the suspension. With this, perhaps wemay be in a better position to intelligently interpretthe results of any spectroscopic measurements on turbidsystems.

SUM MARYThe scattering of light by Chllorclla pyrenoidosa

was investigated by measuring absorption spectra asinfluenced in different ways by scattering. Two seriesof measurements were carried out. In one case thecells were suspended in buffer solution and absorptionspectra were measured with the light detector observ-ing only that part of the total transmitted or scatteredlight coming from the suspension at well defined angles.In the other case, the cells were suspended in mediaof various indices of refraction and absorption spectra

were measured by allowing the light (letector to ob-serve only directly transmitted light. The apparentpositions of the absorption bands were found to bestrong functions of both the angle of observation of thelight detector and the index of refraction of the medi-um. The latter results were found to agree well withthe predictions of scattering theory.

The vessel and baffle assembly in figure 1 wereconstructed by Mr. Richard Hart, who also drew mostof the figures. AMessrs. J. A. MIcClure and C. W.Hamlet performed some of the calculations. \Ve alsowish to thank Drs. C. S. French and J. H. C. Smithand Mr. H. W. Milner for their many helpful sug-gestions. It is appropriate at this time to point outthat our study of light scattering by cell suspensions,which began when the author was a studlent at thePhotosynthesis Project of the University of Illinois,was first suggested by the late Dr. Robert Emerson.

LITERATURE CITED1. BARER, R. Spectrophotometry of clarified cell sus-

pensions. Science 121: 709-715. 1955.2. DITCHBURN, R. WV. Light. P. 459. Interscience,

New York 1954.3. DUYSENs, L. N. M. The flatteninlg of the absorp-

tion spectrum of suspensions, as compared to that ofsolutions. Biochim. Biophys. Acta 19: 1-12. 1956.

4. FRENCH, C. S. Fluorescence spectrophotometry ofphotosynthetic pigments. In: The Luminescenceof Biological Systems, F. H. Johnson, ed. Pp.51-74. A.A.A.S., Washington, D.C. 1955.

5. LATIMER, P. and RABINOW'ITCH, E. Selective scatter-ing of light by pigment containing plant cells. Jour.Chem. Phys. 24: 480. 1956.

6. LATIMER, P. and RABINOWITCH, E. Selective scatter-ing of light by pigment-containing plant cells. In:Research in Photosynthesis, H. Gaffron, ed. Pp.100-107. Interscience, New York 1957.

7. LATIMER, P. Apparent shifts of absorption bandsof cell suspenIsions related to selective light scatter-ing. Science 127: 29-30. 1958.

8. LATIMER, P. Erroneous absorption spectra of livecells caused by anomalous light scattering. In:Proceedings of the National Biophysics Conference,H. Quastler, ed. Yale University Press, NewHaven (In press.)

9. LATIMER, P. and RABINONNITCH, E. Selective scatter-ing of light by pigments in vivo. Arch. Biochem.Biophys. (In press.)

10. LOTHIAN, G. F. and LENNIS, P. C. Spectropho-tometry of granulated materials with particularreference to blood corpuscles. Nature 178: 1342-1343. 1956.

11. SHIBATA, K., BENSON, A. A. and CALVIN, M. Theabsorption of suspensions of living micro-organisms.Biochim. Biophys. Acta 15: 461-470. 1954.

12. VAN DE HULST, H. C. Light Scattering by SmallParticles. Pp. 179-193. Wiley, New York 1957.

199

