It's Chocolate Nemesis, an aptly named dessert that has been the downfall of dozens, maybe even hundreds, of cooks who have tried to make it according to the directions in the best-selling "Rogers Gray Italian Country Cookbook," by Ruth Rogers and Rose Gray. This creation, which is described as a "jelly," is one of the most requested desserts at the authors' River Cafe in London. Kate Dyson, who owns the Dining Room Shop in London, told me that her attempt "looked like a floppy cow pie." So did mine.
For my second attempt, I followed the recipe in the 1996 American edition (both are Random House books), which calls for half a cup more sugar than the English edition does. It worked. Sort of. It did not run all over the plate, but it was as wet as a chocolate sauce.When a London newspaper asked three chefs to make the infamous dessert, all flunked. "It is a sort of challenging cake," Gray acknowledged in a telephone conversation. Dyson says she knows only one person who was successful with it, describing her as "an ace cook who adapts as she goes along and never reads recipes properly."
A recipe that so few people can make successfully is one that cries out for additional testing, not only by the authors but also by others. In that regard, Chocolate Nemesis is not so much an aberration as an example of a problem that bedevils the cookbook industry.
The prevalence of errors in cookbooks is the publishing world's dirty little secret. The problem is likely to get worse as an industry mired in economic doldrums resorts to cost-cutting, practically guaranteeing less editing and testing before publication.
Cookbook editors recognize and, to a degree, expect errors of varying magnitude to pop up and don't consider it a serious problem. But when the cake falls and the mousse doesn't jell, cooks are not so blithe.
Ask Daniele Nugent, a travel agent in Washington, about the Royal Raspberry Tart she made for her parents' wedding anniversary. "I didn't want to make one of those ordinary cakes," she said. "So I followed the recipe from a French pastry cookbook to a T and it was a terrible-tasting, ugly-looking affair. The filling leaked, and it began to burn; the pastry was hard. Thank God we had some sorbet in the freezer."
The problems with the Chocolate Nemesis recipe are numerous. Directions say: "Bake in the oven for 30 minutes or until set. Test by placing the flat of your hand gently on the surface." Should the cake wiggle? Should it be firm? Does it require another 10 minutes, another 20 minutes of baking?
Like any cookbook author challenged about a recipe, Gray offered a number of reasons for the failure. "It's quite tricky to judge people's ovens," she said, "and there's the kind of chocolate you use." But she wasn't sure the difference in the amount of sugar was responsible. That mistake, she said, was the result of someone translating cups into grams and then into pounds and ounces and then back to cups.
Most people would find Gray's solution to the "challenging cake" a trifle annoying: "It's a recipe you need to make a couple of times before you get it right."
If there are so many variables in a recipe, wouldn't more explicit directions and some description be helpful? If the kind of chocolate can make such a difference, shouldn't it be specified? The problems are like so many unexplained mysteries common to cook-books, even the best of them.
Maida Heatter was upfront about one disaster, Lemon Buttermilk Cake. No one could make the version that appeared in the first edition of "Maida Heatter's New Book of Great Desserts" (Alfred A. Knopf, 1982). In a later edition, she offered a revised recipe and wrote in its introduction that the cake in the first edition "sank and was like a wet pudding." She blamed "demons."
Publishers, however, frequently blame authors for these errors. And, apart from Heatter, authors blame publishers (who pay too little, some say, to make the hiring of independent testers feasible).
"Book publishing contracts are very specific," said Sydny Miner, the cookbook editor for Simon & Schuster. "Publishers don't have any way to amortize testing of recipes, so the responsibility falls on the author. We take it on faith that the recipes have been tested."
Some cookbook authors not only test themselves, but also invite culinary students to test for them (gratis); others get their friends to do it.