No abstract

Certain deletion mutants including the tryptophan operon in Salmonella typhimurium are unable to utilize several sugars as carbon sources in solid media, although they are able to grow in liquid media with these sugars. The addition of citrate or washing the agar with ethylenediaminetetraacetic acid permits growth on solid media. Analysis of the agar revealed that Fe'+ and Cr3+ were present at concentrations of 22 and 75 ,UM, respectively. The addition of Fe3+ to liquid media in 0.5 mm concentrations did not inhibit the wild type or the mutants. A similar concentration of Cr3* did not inhibit the wild type, but concentrations as low as 0.01 to 0.05 mm inhibited the deletion mutants. Other metals were inhibitory at various concentrations, but none showed any significant differential effects on the mutants and the wild type. The increased sensitivity of the mutants to chromium may be due either to an increased permeability to Cr3+, resulting in higher effective intracellular concentrations and inhibition of one or more metabolic functions, or to a binding of Cr34 to an altered cell wall, resulting in decreased permeability of required substrates. An investigation of deletion mutations of the

Using the scanning device the thickness of the layer to be removed can be set by simply selecting appropriate values of the pulse-repetition-rate pulse-length product. Although the maximum penetration into the surface being analyzed may be from 1 to 4 pm greater than the mean eroded depth, the volume of material taken from below the eroded layer is relatively small. A layer with a distinct composition 10 pm thick may be analyzed with a sensitivity of ppm atomic without interference from the underlying layer. Errors due to dilution with the underlying layer do not become significant until the surface layer is less than 0.1 pmin thickness.The sensitivity to surface impurities is very high (limit of detection of the order of 10" atoms/cmz) but they can be identified by their rapid decay with successive scans.The technique can be applied to insulators coated with a conducting layer, and thin conducting films on insulating substrates.While the scanning device was designed for the analysis of semiconductor materials, it is obviously applicable in any analysis where a flat surface of sufficient area is available and to the analysis of solutions. ACKNOWLEDGMENTThe authors thank the members of the Engineering Division of the Mullard Research Laboratories concerned with the manufacture of the scanning device and the staff of the Metallurgy Laboratory for carrying out the low angle sectioning. The assistance of the Measurement Department of Mullard Mitcham in carrying out the Talysurf measurements is also acknowledged. Some of the various factors are studied which affect the resulting signal in atomic absorption and ernis-sion-such as changes in fuel and air flow, matrix composition, and aspiration rate-are extremely difficult to control, and will adversely affect the precision. Internal standardization which effectively reduces or eliminates the variations in these factors allows the high precision measurements of which flame spectroscopy is capable to be attained. Some selection rules and experimental studies for determining the efficiency of an internal standard are presented. Limitations in the use of internal standards are discussed. Hybridization techniques-i.e., simultaneous emission and absorption measurements-can be used for greater latitude in the selection of internal standards. Increases of 2-to 3-fold in precision as well as higher accuracy are found utilizing internal standardization.

A method for total calcium in serum was optimized, a dual double-beam atomic absorption spectrophotometer being used, with strontium as an internal reference. The effect of HCl and lanthanum concentrations on the signals of both calcium and strontium was determined. The final analysis was unaffected by known sources of interference in serum at a 50-fold dilution of serum in 50 mmolar HCl containing 10 mmol LaCl3 and 0.12 mmol of SrCl2 per liter. Within-run precision for 30 replicate analyses performed during 30 min gave a relative standard deviation (RSD) of 0.36% for the internal strontium reference mode vs. 0.72% RSD for the single-channel mode. Nine consecutive multiconcentration control pools were measured repeatedly for two years; the median RSD was 2.3%. With meticulous technique, long-term variability for total calcium in serum approached 1%. Accuracy is believed to be limited only by the precision, since systematic bias was not detected from recovery studies and comparisons with results of ashing procedures or oxalate precipitation techniques.