50 Optics & Photonics News ■ March 2003

©Corbis

Or How theCosmos Came to

Draw Itself

“Light Writing”The Invention of

Brian S. Baigrie

March 2003 ■ Optics & Photonics News 511047-6938/03/03/0050/5-$0015.00 © Optical Society of America

(Facing page) Photograph of a daguerreotype camera. A portrait of Louis Jacque MandeDaguerre, taken by Charles Meade in 1841, lies beside the camera. (Above) Photograph ofWilliam Henry Fox Talbot (circa 1865).

©Corbis

In the nineteenth century, scientists began to explore

ways of “fixing” the image thrown by a glass lens. The

daguerreotype appeared in the first half of the century,

followed by less expensive techniques, including the tintype.

Astronomers were among the first to employ the new

imaging techniques: the photoheliograph, a device for taking

telescopic photographs of the sun, was unveiled in 1854.

T hanks to improvements in imaging technology, nineteenth century scientistswere able to learn about the world in ever-greater detail, a fact which placedscientific illustrators under increased pressure to produce more accurate

images. At the same time, perhaps as a result of the growing sense that the demand formore accurate images was beginning to exceed the capability of the illustrator’s hand,scientists began to explore ways of “fixing” the image projected by a glass lens, so thatnature could, in effect, draw itself.

This exploration involved two distinct problems. The first was to prepare amedium that was so sensitive to light that it would allow an image to be impressedupon it. The second was to render the same medium insensitive to further exposure,so that the resulting image could be viewed in light without the second exposure causing it any harm.

most serious drawback was that theimage itself was unique—there was nonegative from which multiple printscould be made.

By the late 1850s, inexpensive tin-types, which produced an image on athin metal plate, had eroded the domi-nance of the daguerreotype. In the firsthalf of the 1860s, soldiers in theAmerican Civil War documented theirexperience on tintypes, new light,durable paper prints. People began tomiss the quality of the fragile, bulkydaguerreotype, but in the face of dozensof paper portraits costing the price ofonly one daguerreotype, the daguerreo-type as a commercial process rapidly disappeared.

The wet negativeThe second solution to the problem oflight writing was worked out by theEnglish amateur scientist William HenryFox Talbot. Talbot started conductingexperiments on paper covered with sil-ver nitrate and other photosensitivechemicals in 1834. He correspondedextensively with John Herschel duringthe course of these experiments. Theprocess that Talbot invented on his ownwas flawed because his negatives andprints faded after a short time. Althoughhe had carefully concealed fromHerschel his original method of fixingpaper to render it insensitive to furtherexposure to light, it was Herschel whoprovided the literal and figurative fixthat Talbot needed. In 1819, Herschelhad discovered that sodium thiosulfate(also referred to as hyposulfite) dissolved

LIGHT WRITING

©CorbisMyers, Library of Congress

52 Optics & Photonics News ■ March 2003

The daguerreotypeDuring the first quarter of the nine-teenth century, the problem of lightwriting (as it came to be known) wastaken on by many scientists and amateurinventors, who experimented withpapers or plates prepared with light-sensitive chemicals. During the 1830s,two different solutions were placedbefore the public.

The first was developed by the Frenchcommercial artist Louis Jacque MandeDaguerre (1787-1851). The daguerreo-type was made on a sheet of silver-platedcopper, which could be inked and thenprinted to produce accurate reproduc-tions of original works or scenes. Thesurface of the copper was polished to amirrorlike brilliance, then rendered lightsensitive by treatment with iodine fumes.The copper plate was then exposed to animage sharply focused by the camera’swell-ground, optically correct lens. Theplate was removed from the camera andtreated with mercury vapors to developthe latent image. Finally, the image wasfixed by removal of the remaining photo-sensitive salts in a bath of hyposulfite(sodium thiosulfate) and toned with goldchloride to improve contrast and dura-bility. Color, made of powdered pigment,was applied directly to the metal surfacewith a finely pointed brush.

Daguerre’s attempt to sell his process(the daguerreotype) through licensingwas not successful, but he found anenthusiastic supporter in FrançoisArago, an eminent member of theAcadémie des Sciences in France. Arago

1839The French governmentcompensatesDaguerre forinventing thedaguerreotype,enabling world-wide use of thetechnique.

1840sOptical meansused to reducedaguerreotypeexposure timesto 3-5 min.

1839-1840John W. Draperphotographs themoon in firstapplication ofdaguerreotypesto astronomy.

Daguerreotype of the moon (1840).

Daguerreotype portrait ofa peddler (1840-1860).

1841William HenryFox Talbotpatents newprocess involvingcreation of papernegatives.

recommended that the French govern-ment compensate Daguerre for his considerable efforts, so that thedaguerreotype process could be placedat the service of the entire world. TheFrench government complied, and theprocess was widely publicized on August 19, 1839, as a gift to the worldfrom France.

The daguerreotype enjoyed animmense following, especially in theUnited States, where fascination with thesilvered plate endured for nearly 20years. The charm of the daguerreotypelay both in its extraordinary beauty andin the fact that it was not hobbled bypatent or license. The formula was freeexcept in England, where Daguerre hadquietly secured a patent, with at least thetacit knowledge of the French government.

The daguerreotype was not withoutits drawbacks. The first cameras requireda lengthy exposure lasting many min-utes. By the 1840s, however, variousoptical means had reduced the exposuretime to three or at most five minutes,and by the end of the decade to a matterof seconds. Still, the image was fixed onhighly polished metal that was heavyand its highly reflective surface madeviewing difficult. Another complicationwas that the metal surface was extremelydelicate, requiring a protective coverglass and a frame or case that made itheavier still. The image itself was limited. The most common size wasapproximately 2.75 in. x 3.25 in. (7 cm x8.2 cm), and it could only be madelarger with great difficulty. Finally, the

silver salts. Herschel suggested thatTalbot try replacing his brine wash andpotassium bromide with hyposulfite ofsoda, which washed off the light-sensi-tive silver-nitrate salts that were respon-sible for the fading images of Talbot’soriginal process. Although Talbot wasslow to accept the new process, he even-tually did and added it to his patentapplication in 1841.

Unlike Daguerre’s crisp images onmetal plates, Talbot’s process (sometimescalled the Calotype or Talbotype) pro-duced paper negatives from which rathersoft, painterly paper prints were made ina separate step. The transparent image,or negative, could be used repeatedly toproduce prints on paper. His discoveryand the process of photogenic drawingwere announced in 1839 to the RoyalSociety. The primary difference betweenthe two processes was that Daguerre’sprocess was considerably quicker and theimages on metal were more detailedthan were Talbot’s. Although Talbot hadinvented photographic paper that wassuperior to any other product in existence, the fibers of the paper elimi-nated the minute details that Daguerre’sprocess revealed with stark clarity.

Acceptance of the Talbotype was alsohobbled by the fact that the process waspatented; commercial licenses were diffi-cult to sell in the face of the essentiallyfree and established daguerreotype. Soalthough Fox Talbot’s prints-from-nega-tives approach attracted a talented fol-lowing and would eventually become thestandard method in photography, in theearly days it was no match for the

March 2003 ■ Optics & Photonics News 53

LIGHT WRITING

daguerreotype in the eyes of portrait-making entrepreneurs and the portrait-buying public. It took two decades ofevolution and new inventions to finallybring about the full acceptance ofnegative–positive photography.

Photographing the skiesAstronomers were among the first toembrace the new photographic techno-logy, contributing greatly to its improve-ment during the nineteenth century.This in itself was not surprising. Attimes, the image of a telescope is notfleeting and remains visible long enoughfor an illustrator to catch its outlines. Ifextremes of heat and cold do not miti-gate against the use of a pencil, theimage can readily be captured by anillustration. These occasions are exceed-ingly rare, however. As often as not, thebriefness of an astronomical phenome-non fatally compromises the prospect ofan illustration. In contrast, the rapidityof response of a photographic platemakes it possible to capture even thebriefest astronomical event in an image.Another considerable advantage of pho-tography is that light much too feeble toexcite vision, if it is given enough time,can impress an image on a sensitive pho-tographic plate. By use of the most powerful telescope, the camera has thecapacity to grasp what is invisible andbring into view a deeper universeheretofore unseen by the most acuteobserver. A ray of light from a star, how-ever feeble, continues to impress the pellicle of a plate, minute after minute,hour after hour, night after night, until

1860Solar prominencesphotographed byWarren de la Rueusing his new photoheliograph.

1879Improvements tophotographic plates allow exposure times asshort as 1/125 of a second.

at last the star, too faint to be detected bya telescope, imprints its image on theplate. Mechanical precision is a prerequi-site for this sort of work. Twenty-fivehours of exposure over the course of tennights to obtain an image is not unusual.

During the winter of 1839-1840, JohnW. Draper, professor of chemistry atNew York University, was the first toapply Daguerre’s process to astronomywhen he succeeded in taking a good picture of the moon. Five years later,William Cranch Bond (1789-1859),director of the Harvard Observatory,took a series of lunar photographs. Withthe assistance of a professional photo-grapher, Bond took the first daguerreo-types of stars in 1850. On April 2, 1845,Armand Hippolyte Fizeau (1819-1896)and Jean Bernard Léon Foucault (1819-1868), who published a number ofpapers that helped to improve thedaguerreotype, took clear images of thesun, which was photographed withincreased success because it was one ofthe objects best suited to the slow speedof photographic material available atthat time. The stars also presented pho-tographers with images that were stableduring the many minutes and evenhours of exposure, but the sun had theadded advantage of high luminosity.

Great strides were made by WarrenDe la Rue (1815-1889), an amateurastronomer. By 1852 he had taken pho-tographs of the moon that could beenlarged without blurring. Ten yearslater, he was producing photographs thatshowed as much as could be seenthrough any telescope. In 1854, he

1850sAmbrotypes, easyto tint and cheaperto make and sell, gradually eclipsethe daguerreotype.

1860sAmerican Civil Wardocumented withtintypes, a newtechnology basedon thin metalplates and paper.

Tintype photograph (1860-1880).A dismantled ambrotype photograph (1860).

Washington CountyHistorical Society,

Minnesota

Washington County Historical Society

54 Optics & Photonics News ■ March 2003

LIGHT WRITING

designed the photoheliograph, a devicefor taking telescopic photographs ofthe sun, which he used in 1860 to takedramatic photographs of solar promi-nences during the total eclipse, provingthat they were indeed solar and not (aswas supposed) lunar in origin.

Most nineteenth century photo-graphs were made on glass-plate nega-tives, except for the Calotype, whichused a paper negative, and thedaguerreotype and tintype, neither ofwhich required negatives. Since earlyglass-plate negatives used a process thatrequired them to be coated just beforeuse, they were known as wet-plate neg-atives. Dry-plate negatives, which didnot require the photographer to makehis own plates, were introduced in1864, but they were so insensitive thatthey did not initially cut into the domi-nance of the glass-plate negative.

A new era of observation opened upin 1879 with improvements to the dryphotographic plate, making it moresensitive—1/125 of a second for properexposure. The higher speed of the newmaterial and the fact that a relativelylong period of time could elapsebetween preparation and use led to therapid diffusion of photographic platesas a means of observation. Almostimmediately, Henry Draper (1837-1882), the son of John Draper, aban-doned the wet-plate process for the dryphotographic plate. He photographedthe nebula of Orion, the first nebula tobe photographed, with a clock-driventelescope and a 140-min exposure. Ayear later he succeeded in taking a pic-ture of its spectrum as well. By 1882,Draper had taken photographs with aprism lens of over one hundred stellarspectra. The application of this methodproved to be extremely valuable in thecompilation of the first general catalogof stellar spectrum types published in1890 by Harvard University.

sequence involved an evolution in thelife of the stars from stage 1 to stage 4.

In 1887, the director of the ParisObservatory proposed an internationalproject dedicated to producing the firstphotographic chart of the heavens. Itinvolved outfitting many of the world’sgreat observatories with new refractorscustomized for photographic purposes.This project had already been antici-pated by Edward Pickering (1846-1919)and his co-workers at HarvardUniversity who, between 1886 and1889, made a complete spectrographicsurvey of the stars of the northernhemisphere.

The most significant result of theuse of photography in astronomicalwork was that it led to the discovery ofnew celestial bodies that might other-wise have remained undiscovered.These discoveries were based on onefundamental fact about stellar photo-graphy, namely, since stars are virtuallymotionless, they register on a photo-graphic plate as tiny round dots.Astronomers hypothesized that, if aphotographic plate were exposed over aperiod of time in a camera anddirected by clockwork to a particularpoint in the sky, many bodies havingappreciable motion across the field ofview (e.g., asteroids and comets) mightregister on the same plate as minutebut nevertheless measurable streaks. Itwas just such a streak that, on August 13, 1898, disclosed the existenceof Eros, an asteroid approximately 10 miles (16 km) in diameter that atthe time was thought to approach earthmore closely than any heavenly bodyexcept for the moon.

Nature at last had drawn itself,demonstrating in the most dramaticway possible the superiority of the newlight writing over traditional methodsof astronomical illustration.Observations of Eros enabledastronomers to compute more precisedistances of the sun and the planets. Inthe same manner, the ninth satellite ofSaturn was discovered by WilliamPickering (1858-1938).

Brian S. Baigrie (baigrie@chass.utoronto.ca) isassociate professor at the Institute for History andPhilosophy of Science and Technology, Universityof Toronto,Toronto, Canada.

Other astronomers soon began tophotograph the spectra of the stars, andit became clear that the stars could begrouped according to the lines in theirspectra. Pietro Angelo Secchi (1818-1878) proposed four spectral classes in1867: class 1 with a strong hydrogenline; class 2 with numerous lines; class3 with absorption bands rather thanlines, which were sharp toward the redand fuzzy toward the violet; and class 4with bands that were sharp toward theviolet and fuzzy toward the red.Secchi’s classes mark a fairly straight-forward temperature sequence. Class 1included blue and white stars; class 2,yellow stars; class 3, orange and redstars; and class 4, deep red stars alone.It was difficult to avoid the conclusionthat the stars of each type differedphysically from those of other types. Ifthe sequence was one of decreasingtemperature from blue and white starsof class 1 to deep red stars of class 4, theconclusion might be drawn that the

Most early photographs were made on glass-platenegatives, except for the Calotype, which used apaper negative, and the daguerreotype and tintype,neither of which required negatives.

George Peter Alexander Healy, one of themost successful portrait painters of themid-nineteenth century. Mathew Brady'sstudio, half plate daguerreotype, between1844 and 1860.

Daguerreotype Collection, Library of Congress