• R Ananthakrishnan

Articles written in Proceedings – Section A

• Photoelectric photometry of light scattering in fluids

The paper deals with a photoelectric method for measurement of the intensities of scattering in liquids, which is much quicker and more convenient than the visual methods hitherto employed. Results of the measurements in the case of a few liquids show good agreement between the results of photoelectric and visual photometry. The possibility of extending the method to depolarisation measurements is outlined.

• On the convergence error in depolarisation measurements

The scattering of light in an interference field is discussed, and it is shown that for the simple case of two parallel plane-polarised intersecting beams, the depolarisation at the interference maxima gives the correct value, while at the interference minima the depolarisation ratio is reversed. The average of the depolarisation taken over the whole field is higher than the correct value. The treatment is extended to the cases in which a lens covered with a square aperture, and with a circular aperture, respectively, is used to concentrate the light on the scattering medium. It is shown that the observed values of the depolarisation would deviate from the genuine values by a correction factor which involves the square of the angle of convergence. The observed depolarisationρ is given by\eqalign{ & \rho _v = \rho _{v_0 } + \alpha \theta ^2 \cr & \rho _h = 1 + {1 \over {\rho _{v_0 } }}\alpha \theta ^2 \cr & \rho _u = \rho _{u_0 } + 2\alpha \theta ^2 \cr} whereα has the value 1/3 for a square aperture, and 1/4 for a circular one. The subscriptsv, h, u refer to the cases in which the incident light has its electric vector vertical, horizontal and is unpolarised respectively. It is pointed out that the same results follow by treating the incident beam as a bundle of incoherent rays. Some consequences of the theoretical results are discussed and the necessity for the perfection of the optical parts used in depolarisation work is empohasised. Experimental results are given which illustrate the points discussed in the paper.

• Redetermination of the depolarisation of light scattering in gases and vapours

It is pointed out that the existing depolarisation data of gases and vapours are gravely defective, and a redetermination of the values has been made with improved experimental technique. The results obtained are strikingly different from those of previous workers, and afford at the same time a natural explanation of many of the existing anomalies. The genuine depolarisation of methane appears to be only of the order of 0·3% which is in all probability to be attributed to the highly depolarised vibration Raman lines. The depolarisation shows a steady increase as one goes to the higher members of the homologous series, but the values in all cases are much smaller than they were hitherto assumed to be. Cyclopropane shows a very small depolarisation in conformity with the known small depolarisation of cyclohexane, while propylene shows a high value as would be expected from the presence of the double bond in it. Methyl chloride is more anisotropic than ethyl chloride, and the depolarisation factor of the former is much higher than that of methane. Carbon tetrachloride shows an extremely small depolarisation of 0·15% which is discussed in detail. It is tentatively concluded that the depolarisation of argon is nil. The depolarisation of H2S is only 0·3%. An explanation is offered for the low depolarisation of the hydride molecules in general.

• The raman spectrum of heavy water

• Some new features in the raman spectra of carbon and silicon tetrachlorides

The Raman spectra of CCl4 and SiCl4 have been examined under low and high dispersion. In both cases a faint line has been observed close to the line corresponding to the total symmetric vibration. This faint line has a frequency shift of 434 cm.-1 in the case of CCl4 and is interpreted as the octave of the fundamental frequency 217·9 cm.-1 The corresponding line in the case of SiCl4 shows a frequency shift of 440·5 cm.-1 and is interpreted as the octave of the fundamental frequency 221 cm.-1 It is pointed out that the appearance of the octave in these cases might probably be due to its proximity to the total-symmetric vibration frequency which is respectively 459 cm.-1 and 423 cm.-1 in the case of CCl4 and SiCl4. An additional faint line with a frequency shift of 145 cm.-1 has been observed in the Raman spectrum of CC14 which is interpreted as a differential tone arising from the fundamental frequencies 459 cm.-1 and 314 cm.-1, the summational of which causes the splitting of the fourth frequency, and the octave of whose summational appears as a faint broad line at 1538 cm.-1 The broadening of the fourth frequency in the case of SiCl4 might be due to a cause similar to that which causes the splitting of the fourth frequency in the case of CCl4. The total symmetric vibration in the case of CCl4 shows a well-defined wing to its red side, sharply separated from the principal line.

• The Raman spectra of some organic liquids under high dispersion and resolving power - Benzene, toluene, phenol, chlorobenzene, pyridine and cyclohexane

• Polarisation of the raman bands of water and deuterium oxide

The Raman spectra of liquid H2O and D2O have been photographed by interposing a double image prism in front of the slit of the spectrograph. The three components of the principal Raman band in the two cases are more and more depolarised in the order of increasing frequency, the component of lowest frequency being highly polarised, and that of highest frequency being highly depolarised. The band Δv = 1235 cm.-1 in the case of D2O is imperfectly polarised, while the low frequency Raman band (Δv≈ 175 cm.-1) is depolarised to the extent of 6/7 in both cases. A tentative explanation is suggested to interpret the observed results.

• Polarisation of the raman bands of water and deuterium oxide

The Raman spectra of liquid H2O and D2O have been photographed by interposing a double image prism in front of the slit of the spectrograph. The three components of the principal Raman band in the two cases are more and more depolarised in the order of increasing frequency, the component of lowest frequency being highly polarised, and that of highest frequency being highly depolarised. The band Δv=1235 cm.−1 in the case of D2O is imperfectly polarised, while the low frequency Raman band (Δv≈175 cm.−1) is depolarised to the extent of 6/7 in both cases. A tentative explanation is suggested to interpret the observed results.

• The Raman spectra of propylene and ISO-Butane

The Raman spectrum of propylene has been reported in detail and the observed frequencies are discussed in the light of the knowledge of their state of polarisation. The frequencies 1648, 920 and 432 are identified as the fundamental frequencies of the carbon chain, the first two being of the valence type, while the third one represents a deformation frequency. The fourteen additional lines are to be attributed to the oscillations of the hydrogen atoms. The Raman lines of propylene are characterised by extreme sharpness, and are more or less well polarised. The Raman spectrum of iso-butane forms a striking contrast, many of the lines being extremely broad and highly depolarised. It is suggestd that this difference might be due to the symmetry of the iso-butane molecule which causes many of the vibrations to be degenerate. An attempt has been made to interpret the spectrum of iso-butane with the help of the polarisation data. The close similarity between the Raman spectra of iso-butane and chloroform as revealed by polarisation studies is pointed out.

• The Raman spectra of some boron compounds - Methyl borate, ethyl borate, boron tri-bromide and boric acid

The Raman spectra of methyl and ethyl borates have been carefully investigated, and it is found that the Raman frequencies previously reported for these substances require considerable revision. The Raman spectrum of methyl borate in the region of low frequencies shows great similarity to the spectra of the tri-chlorides of phosphorus and arsenic, and this similarity is strikingly brought out by the polarisation characters of the Raman lines. A pyramidal structure is postulated for the B(OCH3)3 molecule, the group (OCH3) being treated as a single unit. The observed and calculated frequencies are in fair agreement. The Raman spectrum of boron tri-bromide is reported for the first time, and consists of two very intense lines and three very weak ones. These are compared with the frequencies of the tri-chloride. Only one line was found in the case of boric acid, which is attributed to the totally symmetrical vibration of the molecule.

• The Raman spectra of cyclo-propane and ethylene oxide

The Raman spectrum of cyclo-propane has been investigated at the room temperature both in liquid and vapour states. Ten frequencies have been recorded in the spectrum of the liquid of which four are reported for the first time. The state of polarisation of the vibrational Raman lines has also been studied. The four intense Raman lines of the liquid (863, 1187, 3008 and 3028) have also been recorded in the spectrum of the vapour. The significance of the presence of the frequency Δv=863 in the spectrum of the vapour is pointed out. The relative intensity of the Raman lines 3008 and 3028 is considerably modified in passing from the liquid to the vapour state. The Raman spectrum of ethylene oxide also shows ten frequencies of which two have not been previously recorded. The spectrum shows great similarity to that of cyclo-propane. The polarisation characters of the Raman lines have been determined, and the molecular configuration discussed.

• The raman spectra of tri-methyl-amine and some compounds of hydroxylamine and hydrazine

• The Raman spectra of crystal powders - I. The halides and sulphate of ammonium

The paper describes a new technique which has been developed by the author for photographing the Raman spectra of crystal powders. It consists essentially in employing a fairly monochromatic beam for excitation, and absorbing the exciting line from the scattered light before the latter enters the spectrograph, and involves thus the use of a pair of complementary filters. The new technique enables for the first time to obtain Raman spectrograms of crystal powders which are as elegant and rich in detail as those of liquids. The Raman spectra of NH4F, NH4Cl, NH4Br, NH4I and (NH4)2SO4 in the crystalline state at room temperature have been examined by this method. The Raman spectrum of the ammonium ion shows a pronounced dependence upon the crystal structure of the salt as well as upon the nature of anion. The spectra of crystalline NH4Cl and NH4Br at room temperature are particularly intense, and comprise about six distinct frequencies. These crystals are isomorphous at this temperature and there is a striking general resemblance between their spectra in spite of individual differences. All the Raman frequencies of the ammonium ion in these two substances have their analogues in the infra-red absorption spectra. This result is very significant and although it suggestsa priori the lack of tetrahedral symmetry for the [NH4]+ ion, it is pointed out that the influence of the crystalline environment on the vibrations of the ion has to be carefully considered before any definite conclusions are drawn. The spectra of NH4I and NH4F are weaker than and differ from those of the other two halides. This might be due partly to the fact that their crystal structure at room temperature is different from that of the other two halides. The Raman spectrum of the ammonium ion is extremely feeble in the case of (NH4)2SO4 which shows the enormous influence of the complex anion on the vibrations of [NH4]+.

• The Raman spectra of crystal powders - II. The chlorides and sulphates of hydroxylamine and hydrazine

The paper reports the Raman spectra of crystalline NH2OH·HCl, (NH2OH)2·H2SO4; N2H4·2HCl, and N2H4·H2SO4, photographed by the powder technique recently described by the author. The spectra of the chlorides reveal many interesting features relating to the vibrations of the hydroxylammonium and hydrazinium ions, the most characteristic of which is perhaps the multiplicity and diffuseness of the N−H bands whose frequencies range roughly from 2500 to 3200 wave numbers. The N−H bands are considerably weakened and are very much more diffuse in the case of the sulphates. There is a probable structural similarity between the [H3N.OH]+ ion and H3C·OH and [H3N·NH3]++ and H3C·CH3 although the spectra of the ions are very much more complicated than those of the corresponding organic molecules. From a comparison of the Raman spectra of tri- and pentavalent nitrogen derivarives, it is found that there is a definite lowering of the N−H frequencies in the compounds of the latter class. It is therefore concluded that when the covalency of nitrogen changes from three to four, there is a definite weakening of the N−H bond.

• The Raman spectra of crystal powders - III. Exchange reactions: NH4Cl and D2O

The paper deals with the application of the Raman effect to the study of the exchange reaction between NH4Cl and D2O. The Raman spectra of thed-ammonium chlorides ND4Cl, ND3HCl and ND2H2Cl obtained by exchange reactions have been reported. The spectral data are complete in the case of ND4Cl while they are incomplete in the case of the other two compounds. The Raman spectrum of ND4Cl is compared with that of NH4Cl and discussed in some detail. The spectrum of the water recovered from the exchange reaction has been photographed directly and also as water of crystallisation in strontium chloride. The results are compared with the Raman spectra of liquid H2O and D2O in the one case, and with the spectrum of SrCl2·6H2O in the other. There appears to exist differences in the structure of the water band in the Raman spectrum of the water of crystallisation of the salt when H2O is replaced by D2O.

• The Raman spectra of crystal powders - IV. Some organic and inorganic compounds

• The Raman spectra of some simple molecules - Dimethyl ether, phosgene,n-butane, ethylene diamine, ethylene glycol, ethylene dichloride, ethylene dibromide, acetylene tetrachloride, acetylene tetrabromide and hexachloroethane

The Raman spectra of dimethyl ether, phosgene,n-butane, ethylene diamine, ethylene glycol, ethylene dichloride, ethylene dibromide, acetyleue tetrachloride, acetylene tetrabromide and hexachloroethane have been studied afresh. The Raman frequencies of phosgene have been classified with the aid of polarisation data. Polarisation measurements have also been made in the case ofn-butane, ethylene glycol, ethylene dichloride, ethylene dibromide, and acetylene tetrabromide. A brief discussion of the spectra of the ethane derivatives in relation to the problem of “free rotation” is added.

• Effect of temperature on the Raman spectrum of liquid CCl4 - Some preliminary considerations

• Evidence for the existence of the “emission layer” in the atmosphere

The theory of radiative equilibrium demands that on the average the total amount of energy absorbed by the earth and its atmosphere in the form of short-wave solar radiation should be exactly equal to the total amount of energy given back to space in the form of long-wave heat radiation. From a study of the absorbing and radiating properties of the atmosphere, F. Albrecht arrived at the fundamental result that the major contribution to the long-wave heat radiation into outer space originates from a layer of some three to four kilometres thickness in the upper troposphere, which he designates as the “Emission Layer”. The emission layer is thus a portion of the upper atmosphere which is continually cooling due to radiative loss of heat. The height of the emission layer is a function of the water vapour content of the atmosphere; it is more when the atmosphere is hot and humid and less when the atmosphere is cold and dry.

The author has made a detailed study of the thermal structure of the atmosphere over Agra based on the results of sounding balloon ascents over a period of ten years. A number of interesting features find a ready explanation on the assumption that the emission layer over Agra is located approximately between 11 and 14 gkm. in the monsoon months and between 8 and 11 gkm. during the remaining months, —an assumption in conformity with Albrecht’s work. The observed seasonal variations in the thermal structure of the atmosphere over Agra thus lend, strong evidence for the existence of the emission layer in the atmosphere and the variation of its altitude depending upon the moisture content of the atmosphere.

• Erratum to: On fluctuations of pressure and temperature in the atmosphere

• On fluctuations of pressure and temperature in the atmosphere

The paper contains a critical survey of the problem of pressure and temperature fluctuations in the atmosphere. If we exclude non-adiabatic processes, then changes of pressure and temperature are essentially due to horizontal and vertical movements in the atmosphere. The vertical distribution of pressure over the globe shows that horizontal movements (advection) can give rise to large changes of pressure especially in the middle latitudes. The problem of identifying the layer or layers in which addition or removal of mass has taken place due to advection, from the data furnished by aerological ascents over a station has been tackled by a number of investigators. Rossby considered the general problem of advection in several layers of an atmospheric air column, but his theory when applied to practical cases yields results which are often unreasonable. Ertel and Sjan-zsi-Li considered the same problem some years later and claimed to have discovered a fundamental error in Rossby’s theory. In the present paper this problem is considered at some length. It is shown that Rossby’s theory is quite correct under the assumption on which it is built up while the theory of Ertel and Li is not valid for cases of arbitrary advection. It is shown that the neglect of horizontal advection of pressure and horizontal convergence and divergence of wind velocity is responsible for the obviously unreasonable results furnished by Rossby’s theory when applied to actual examples.

• Erratum to: On fluctuations of pressure and temperature in the atmosphere

• Spectrophotometric study of sunspots

After a brief review of some of the earlier work on sunspots the experimental technique employed for the spectrophotometry of sunspots and the results obtained in the case of 5 spots are described. In the case of one of the spots studied on 1951 May 13, the intensity ratio umbra/photosphere decreased from about 32% at λ 6620 A to 15% at λ 5078 A. The lowest value of 22% at λ 6620 A was recorded for the umbral intensity of a spot on May 24. Microphotometer records of two spots are reproduced showing the discontinuous nature of the transition from the umbra to the penumbra first observed visually by Secchi. Tables and curves of I. (λ, T*)/I (λ, T) based on Planck’s radiation law are given and the effective temperatures of the spots calculated. Odgers’ theory of the “bright rings” in sunspots is considered and it is pointed out that the observed small values of excess intensity of the bright rings probably imply that sunspots extend to appreciable depths below the photosphere.

• Errata

• Prominence activity (1905–1952)

The paper illustrates with a series of diagrams the variation of prominence activity during the period 1905 to 1952 based on the prominence statistics collected at the Kodaikanal Observatory. The main features of prominence activity brought out by these diagrams are briefly discussed.

•